Review article

MANAGEMENT OF HERNIATION OF THE LUMBAR DISC F. Postacchini

The natural history of lumbar disc herniation

Clinical evolution. It is extremely difficult to study clinically the natural history of a condition causing pain, since patients almost inevitably undergo some form of treatment. This may explain the paucity of information on the natural evolution of the clinical symptoms and signs of disc herniation. In a multicentre prospective study, Weber, Holme and 1 Amlie analysed 208 patients who presented with the clinical features of lumbar radiculopathy probably due to disc herniation. In no case was herniation diagnosed by imaging studies. All the patients were examined between two and four weeks after the onset of symptoms and a questionnaire was used to evaluate their clinical status at three and 12 months. All were instructed to observe complete bed rest for one week. Some were treated with piroxicam, whereas the others were given a placebo. No signicant difference in the evolution of signs and symptoms was observed between the two groups. During the rst four weeks after the onset of the symptoms, 70% of patients had a considerable decrease in pain and almost 60% had resumed work. By one year, some 30% complained of back pain, decreased working ability and limitation in recreational activities; 19.5% had not resumed work. Four patients had been treated by operation. In a prospective, randomised double-blind study by Fras2 er, 30 patients had chymopapain chemonucleolysis and 30 were injected with saline. Disc herniation was diagnosed by myelography in all patients. At six weeks, only 37% of patients in the placebo group had a satisfactory clinical result. This increased to 57% at six months but had 3 decreased to 47% by two years. Operation had been undertaken in 40% of patients. The results of this study are 1 not consistent with those reported by Weber et al, but, in the latter trial, the diagnosis of disc herniation had been made solely on clinical grounds by non-specialists. Fras2 ers patients had been referred to a specialist centre, probably after failure of conservative management, and the
F. Postacchini, MD, Professor of Orthopaedic Surgery Clinica Ortopedica University of Rome La Sapienza, Piazzale Aldo Moro 5, 00185 Rome, Italy. 1999 British Editorial Society of Bone and Joint Surgery 0301-620X/99/410213 $2.00 J Bone Joint Surg [Br] 1999;81-B:567-76.
VOL. 81-B, NO. 4, JULY 1999

clinical diagnosis of disc herniation had been conrmed by myelography. Pathomorphological evolution. In recent years, numerous 4-7 studies have shown that a disc herniation may decrease in size or disappear in the course of a few months, no matter whether it is contained, extruded or migrated, or of 4 a small or large size (Fig. 1). In a prospective study, 111 patients with disc herniation or annular bulging diagnosed by CT, had a second CT one year later after one or more epidural injections of steroids. Of the patients with disc herniation, 76% showed a decrease in size, with one-fth of those demonstrating disappearance of the protrusion, on control CT scans. Only 29% of patients with a bulging annulus brosus showed such shrinkage. Deterioration was observed on CT scans in only four patients (5%). Similar 6 ndings were observed by Maigne et al ; of 48 patients who had a further CT scan one to 48 months after the initial examination, 64% showed a decrease of over 75% in the size of the herniation with shrinkage of between 50% and 75% in 17% of the cases. Large herniations tend to decrease in size to a greater 5-7 extent, but extruded protrusions of small size show less tendency to spontaneous resolution. A decrease in size may occur in the course of a few weeks before complete resolu8 tion of the symptoms. A retrospective study has shown that after a mean period of 262 days, most extruded herniations had become smaller or had disappeared after conservative management, but few of the contained protrusions showed any signicant change. Little is known about the mechanisms leading to these changes. In contained protrusions, the main mechanism is likely to be dehydration of the herniated nucleus pulposus. This may account for the higher frequency with which young subjects present a decrease in size of their hernia4 tion. In extruded or migrated discs, phagocytosis of herniated tissue by macrophages probably plays the primary role.

Results of conservative treatment

In a large proportion of patients conservative treatment relieves pain in a few days to several months. Resolution of symptoms may occur in the presence of herniations of any type or size.
567

568

F. POSTACCHINI

Fig. 1a

Fig. 1b

Spontaneous disappearance of L4/L5 disc herniation in a patient with herniation at the two lowermost lumbar discs. Figure 1a Sagittal MRI showing disc herniation at the L4/L5 and L5/ S1 levels. The L5/S1 herniation was responsible for severe compression of the left S1 nerve root; surgery was performed at this level with complete resolution of symptoms, whereas the L4/L5 disc herniation (arrowhead) was not excised. Figure 1b MR image obtained 13 months after surgery. The L4/L5 disc herniation has disappeared (arrowhead).

In a retrospective study, 58 patients with disc herniation treated conservatively by analgesics, anti-inammatory medication (NSAIDs), epidural injection of steroids, at a low back school or by exercises, were followed for a mean period of 31 months. Surgery was necessary in 10% because of inadequate resolution of the symptoms. Of the remaining 52, 50 had an excellent or good clinical result and 48 resumed work after a mean period of 3.8 months. Of the patients with extruded herniation (26%), 87% obtained satisfactory results and all returned to work irrespective of the presence of a neurological decit. In another series of 4 114 cases, 14% of patients required surgery because of failure of conservative treatment; in the remainder, nonsurgical treatment, such as epidural or paraforaminal injections of anaesthetics and steroids, led to satisfactory results. Similar outcomes with 90% of satisfactory results have 6 been reported by Maigne et al using a combination of several treatments such as bed rest, NSAIDs, epidural injections of steroids or corsets. 10 Conservative versus surgical management. Hakelius analysed retrospectively 417 patients treated conservatively by bed rest, a corset and physiotherapy and 166 who had surgery. The patients were assessed monthly for the rst six months after the beginning of treatment or operation and most were followed for a mean of 7.4 years. In the rst month, 76% of patients managed conservatively had beneted from treatment, compared with 97% of the operated patients, but at six months the percentages were similar (93% and 99%, respectively). The mean time away from work was only slightly longer in patients treated conservatively. At six months, the percentage of patients still away from work was 37% in the group with disc herniation demonstrated by myelography and treated conservatively and 7% in the surgical group. In the long term the results

were only slightly better in the patients treated surgically. The incidence of recurrences of radicular pain in the years after conservative treatment was 20% compared with 10% in patients undergoing operation. The limitation of this study is that the choice of treatment was not randomised and thus the two groups of patients are not comparable. 11 In a prospective study by Weber, 280 patients with disc herniation demonstrated by myelography were assigned to three groups. Group I included 87 with mild symptoms who were treated conservatively. The 67 patients in group II in whom there were absolute indications for surgery, had an operation. The 126 patients in group III in whom the need for operation was not so obvious were randomly assigned to conservative (81 cases) or surgical (73 cases) management. All patients in group III were followed up for one, four and ten years after treatment. At one year the percentage of satisfactory results was signicantly lower in the conservative (61%) than in the surgical (80%) group. After four years it was still lower in the conservative group, but the difference was no longer statistically signicant. Comparable results were observed at ten years. Of the 66 patients in the conservative group, 25% had had surgery during the rst year because of the persistence or worsening of symptoms. Neurological decits improved or disappeared in comparable proportions in the two treatment groups. The main defect of this study is that only the patients with uncertain indications for surgery were randomised for treatment. 12 A recent investigation evaluated retrospectively 55 truck drivers, 30 of whom had had prolonged conservative management and 25 an operation. The results of treatment were analysed, as was the cost of health care in the ve years after initial presentation. In both groups, 80% of patients had a satisfactory outcome. There were no sigTHE JOURNAL OF BONE AND JOINT SURGERY

MANAGEMENT OF HERNIATION OF THE LUMBAR DISC

569

Table I. Recent prospective, randomised double-blind studies on chymopapain Fraser Year Number of patients Mean age (yr) Dose (mg) Placebo Follow-up (mth) Success chymopapain (%) Success placebo (%) * chymopapain placebo cysteine-edetate-iothalamate Table II. Long-term and very long-term results of chemonucleolysis with chymopapain Author/s Dubuc et al 24 Sutton 19 Jabaay 15 Dabezies et al 22 Nordby 25 Thomas et al 20 Maciunas and Onofrio 21 Manseld et al 17 Flanagan and Smith 18 Gogan and Fraser 23 Postacchini and Perugia 26 Wilson and Mulholland
16 3

Javid et al

14

Dabezies et al

13

1982 60 (30*, 30) 37.1*, 37.2 8 Saline 1.5 to 6 80 57

1983 108 (55*, 53) 37.9*, 39.9 8 Saline 1.5 to 6 73 42

1988 159 (78*, 81) 37.2*, 38.7 8 CEI 1.5 to 6 71 45

Table III. Results of percutaneous automated nucleotomy in ten clinical series Success rate High 30 Davis and Onik 32 Onik et al 28 Bocchi et al 29 Bonaldi et al 31 Gill and Blumenthal Low 35 Kahanovitz et al 36 Revel et al 33 Dullerud et al 34 Grevitt et al 37 Shapiro Number of patients 200 495 500 237 109 38 69 142 115 57 Follow-up (mth) 6 12 6 11 15 16 12 21 55 27 Success (%) 77.5 66.4 71.0 75.0 79.0 55.0 37.0 56.0 45.0 58.0

Number of patients 842 208 130 94 739 42 268 146 357 30 68 200

Follow-up (yr) 5 6 8 8 8 9 10 10 10 10 5 5 to to to to to to 12 11 10 12 13 13

Successes (%) 81.0 79.0 71.5 80.6 76.0 81.0 80.1 66.0 74.0 80.0 82.0 71.0

(minimum) to 29 to 40 to 60

to 14 to 20 to 10 to 13

nicant differences in the costs of treatment between the two. In this study, only patients with uncertain indications for conservative or surgical management were included.

Percutaneous treatment
Chemonucleolysis. Chemonucleolysis with chymopapain is a technically simple procedure for the L4/L5 and L5/S1 discs and has given the highest rate of satisfactory results of all the percutaneous procedures. In the most recent randomised, 3,13,14 the mean rate of success in the double-blind studies, short term was 74% with chymopapain and 48% with a 15-26 placebo (Table I), and in 12 retrospective studies in which the long-term results were assessed, a successful outcome was achieved in 77% (Table II). The high therapeutic efficacy of nucleolysis is probably due to the enzyme being carried in a liquid agent which is able to reach any area of the disc into which the injection solution can penetrate. Compared with the other percutaneous procedures chemonucleolysis has a higher risk of severe complications, particularly when used indiscriminately in inexperienced hands. Experienced physicians, however, have never reported serious neurological complications or anaphylactic reactions leaving permanent sequelae. None the less, nucleolysis with chymopapain should not be considered as a minor therapeutic procedure representing the last stage of conservative management, but as having clear-cut indications when it is performed on account of its intrinsic advantages. Chemonucleolysis requires careful selection of the patient. Good candidates are those presenting with a small
VOL. 81-B, NO. 4, JULY 1999

or medium-sized herniation, mild or moderate neurological decit, no marked narrowing of the disc, radicular symptoms of less than eight months duration, and no evidence of nerve-root canal stenosis. This does not imply that patients with a large contained herniation, subligamentous extrusion or severe radicular decits may not also have satisfactory results, but in these cases the chances of success are considerably less. The procedure then becomes, to a certain extent, a therapeutic attempt merely aimed at avoiding surgical treatment. Those undergoing chemonucleolysis have similar chances of recurrence of the herniation compared with patients submitted to surgery. Collagenase does not offer any signicant advantages compared with chymopapain. The absence of major allergic reactions is balanced by a lower therapeutic efcacy and a 27 comparable or higher rate of neurological complications. Enzymes still under investigation, such as chondroitinase ABC, cathepsins B and G and calpain I, do not appear to compete with chymopapain. Percutaneous automated nucleotomy (PAN). This is a simple technique at the L4/L5 and more cranial levels, but the L5/S1 disc may be difcult or impossible to approach. Infection of the disc is the only real complication, although, exceptionally, neurological damage has been reported. The ease of the technique and the low rate of complications made the procedure very attractive, until serious doubts arose concerning its therapeutic efcacy. The proportion of 28-32 satisfactory results was high in some studies, but did not improve on those obtained with an intradiscal placebo 33-37 in other series (Tables I and III). This suggests that

570

F. POSTACCHINI

PAN may not be truly effective, the successful outcomes being due in many cases to spontaneous resolution of the symptoms. Studies using serial CT showed that after a mean period of six months the size of the herniation was not modied or had increased in some 75% of patients 38 submitted to PAN. The indications for this form of treatment are so limited, however, that only a small proportion of patients with disc herniation are good candidates for it and in these patients conservative management has a good chance of relieving the symptoms. Little is known concerning the mechanism of this technique and the few available studies suggest that PAN may increase rather than reduce the bulging of the disc in the spinal canal. Manual percutaneous discectomy. This includes the technique performed without the use of an endoscope, percutaneous discectomy (PD), and that carried out with 39-42 endoscopic control (PED). PD usually allows removal of the nucleus pulposus to a similar extent as PED and the results are comparable with those obtained by endoscopic discectomy. The endoscope, however, enables the operator to check the completeness of the discectomy, particularly in the posterior portion of the disc and should be preferred. If strict scientic criteria are applied in the evaluation of the results of manual percutaneous discectomy, the therapeutic efcacy of this method remains to be shown. The number of patients assessed under prospective, randomised and controlled conditions is too small to draw denite conclusions. A few clinical trials support the impression that removal of the nucleus pulposus under endoscopic control can lead to a clinical success rate of about 70% in patients with a contained or small extruded herniation, but if one considers that patients in these circumstances often undergo spontaneous resolution of the clinical symptoms, PD or PED seems to be an adequate procedure in less than 15% of patients needing surgery. Laser discectomy. Numerous experimental studies indicate that various laser systems are able to coagulate, shrink, carbonise, vaporise or ablate the nucleus pulposus, but only a few have been used for clinical purposes. In endoscopic disc surgery, the laser, if correctly used, appears to be as safe as manual instruments with no complications related to its use reported so far. Flexible forceps for manual discectomy, however, are as effective as the laser in the removal of the posterior portion of the nucleus pulposus. Moreover, the use of a laser does not reduce the operating time and is not technically simpler but the cost is considerably higher. The clinical results appear comparable with those obtained with manual or automated 43 percutaneous discectomy. At present, the laser appears to be a tool that is neither necessary nor particularly useful in endoscopic disc surgery, and this may explain why so far its use has been very limited. Conclusions. In the 1980s, there was an explosion of interest in percutaneous techniques, but in the last few years, this has decreased considerably since it has been

recognised that most of these procedures give a proportion of satisfactory results which is only slightly higher than that obtained with conservative treatment or no treatment. Only chymopapain chemonucleolysis continues to have a good reputation in terms of clinical results, but its complications and the advent of microdiscectomy have led to a progressive decrease in the popularity of the procedure. Percutaneous procedures, particularly chemonucleolysis, still have a role in the treatment of a limited proportion of patients with lumbar disc herniation, provided that the indications are based on strict criteria for selection and that they are carried out by experienced surgeons in patients who accept that the chances of success do not exceed 80%.

Surgical treatment
Indications. The indications are absolute in those rare patients with a cauda equina syndrome and in the presence of severe motor decits of recent onset and/or intractable pain. In patients with a cauda equina syndrome, surgical management should always be performed early to increase 44-46 It is the chances of satisfactory neurological recovery. also necessary in the presence of severe sensory and motor decits if the type and size of herniation make spontaneous regression of the symptoms unlikely. In patients with intractable radicular pain, conservative management should be attempted but abandoned if it appears to be ineffective. In all other cases, the indications for operation are relative and depend on four factors: 1) The duration of the radicular symptoms. The chances of resolving symptoms with conservative care decrease progressively with increasing time. After three months of continuous or almost continuous lumboradicular pain, the chances of improvement are slight and decrease further after six months. 2) The type and size of the herniation. It is more likely that the symptoms will decrease in severity or disappear when the herniation is contained and small than in the presence of a large extruded or migrated fragment of disc. 3) The presence of stenosis of the nerve-root canal or the central spinal canal. The neural structures may escape compression by a herniated disc less easily in the presence of a decreased reserve space in the spinal canal, as occurs when the latter is stenotic. 4) The quality and severity of symptoms. Surgery is more often indicated in patients with severe, exclusively radicular, pain than in those with moderate low back and leg pain, since in the former the symptoms are less likely to resolve spontaneously and the results of surgery tend to be better. The presence of a mild or moderate motor decit does not necessarily affect the indication for surgery or conservative management. Surgery should be performed in all patients with a relative indication when no signicant improvement has been obtained with conservative care. The duration of the
THE JOURNAL OF BONE AND JOINT SURGERY

MANAGEMENT OF HERNIATION OF THE LUMBAR DISC

571

Fig. 2a

Fig. 2b

Microdiscectomy. Figure 2a Operative eld at the end of discectomy. The arrows indicate the emptied disc and the asterisk the emerging nerve root. Figure 2b The epidural fat layer above the emerging nerve root (asterisks) is clearly visible.

latter is not well dened but should rarely be less than two months, since it is in this interval that an improvement in symptoms usually occurs. Patients who do not improve considerably after this period have fewer chances of achieving an adequate resolution of symptoms with increasing time. Contraindications. The only absolute contraindication is a disc herniation discovered incidentally in asymptomatic subjects. The other contraindications are relative. Discectomy is generally contraindicated in five situations: 1) When the only clinical abnormality is a mild or moderate motor loss. Even when weakness is severe, however, surgery is rarely indicated. The same considerations apply to sensory decits, which usually disappear spontaneously with time. 2) In patients with psychological disorders or involved in legal controversy, unless a clear-cut organic pathological condition is present. Even then the result may not be satisfactory. 3) In bulging of the annulus brosus. This rarely requires discectomy unless a narrow or stenotic spinal canal contributes to cause severe nerve-root compression. 4) In the presence of vague radicular symptoms, or symptoms in a different dermatome than expected. Based on the level of herniation further investigations should be performed and discectomy should be considered with caution. 5) In patients whose radiated pain is conned to the buttock, who are usually not good candidates for discectomy. It is rarely indicated in patients complaining only of low back pain, in whom fusion of the motion segment should be considered if operation is undertaken.

tion with discectomy or when, in the presence of lumbar stenosis, bilateral laminectomy is carried out before excision of the disc. It may be the procedure of choice in patients with disc herniation at multiple levels or after a recurrent prolapse. It may then be necessary to start the operation under normal vision and use the operating microscope only when the neural structures are visualised. This allows overall assessment of the eld of operation. The conventional operation does not necessarily imply a large exposure with a major laminoarthrectomy.

Microdiscectomy
Use of the microscope. The main drawback of discectomy under normal vision is the lack of adequate illumination of the deep surgical eld. The degree of lighting is related to the extent of exposure of the spine, but even an extensive exposure may not ensure adequate illumination of the spinal canal. The surgeon may be forced to operate without adequate visualisation of the anatomical structures. The operating microscope ensures excellent lighting, regardless of the extent of the surgical exposure and the depth of the anatomical structures. It gives a magnied view, allowing the deep structures to be seen clearly (Fig. 2). The operation may then be performed with greater precision, the causes of compression of the neural structures may be more easily identied and there are fewer risks of causing undue trauma to the emerging nerve root or thecal sac. Use of the microscope allows limitation of both the surgical approach and the extent of the laminoarthrectomy. There are no unequivocal data as to whether better clinical results are obtained after microdiscectomy compared with the conventional procedure. A more limited cutaneous, fascial and muscular access, however, involves fewer risks of complications in healing and less local pain 47-49 The arthrectomy in the early postoperative period.

Conventional discectomy
Conventional discectomy is indicated in any patient with a herniated disc. It is also the current procedure when an arthrodesis of the motion segment is performed in associaVOL. 81-B, NO. 4, JULY 1999

572

F. POSTACCHINI

Fig. 3a

Fig. 3b

MR image and postoperative radiograph of a patient who had a microdiscectomy for a migrated herniation at the L3/L4 level through an interlaminar approach. Figure 3a MR axial scan showing a disc fragment migrated into the proximal portion of the right L3/L4 intervertebral foramen (arrowhead). Figure 3b Postoperative radiograph showing extension of laminotomy (arrows) and preservation of the pars interarticularis.

preserves better vertebral stability and tends to decrease the severity of postoperative low back pain. Patients undergoing microdiscectomy are able to resume their everyday activities more rapidly. These advantages are not such as to make the use of the microscope indispensable, but they do facilitate the work of the surgeon and may improve the quality of the result. This may explain why microdiscectomy has become a popular 47,49-55 technique. Indications. Microdiscectomy is indicated in all patients with a herniated disc at a single level. Use of the microscope is also indicated in recurrent disc herniation, provided that the surgeon has adequate experience in microsurgery. Otherwise, it should be used only at the time of excision of the disc. The presence of stenosis of the root canal in association with a herniated disc at the same level does not affect the indications for the use of the micro52 scope. In patients with two-level herniations the surgical exposure is necessarily extensive and allows good lighting of the surgical eld. In these cases the microscope enables 52 the surgeon to operate with greater precision.

Results of surgical treatment

After operation, the results are satisfactory in two to three months in approximately 85% of patients. In the short term, the percentage of satisfactory results ranges in most studies 26 from 75% to 95%. The results in the medium term are similar. In the long term, the outcome after operation tends to deteriorate in a limited number of cases because of recurrence of radicular pain or exacerbation or recurrence of low back pain. By then, approximately 10% of patients have undergone further operation at the same or different levels to the previous procedure. In the very long term less than half of the patients are asymptomatic. The remainder complain of some symptoms, usually in the low back, the presence and severity of which seem to be related to

degenerative changes independent of the operation. With microdiscectomy, the stay in hospital after opera56 tion may be 24 hours or less. Reduced back pain allows a more rapid functional recovery and a faster return to sedentary work. After eight to 12 weeks, the results of microdiscectomy are similar to those of conventional surgery, provided the arthrectomy is only slightly wider than that currently performed when using the microscope. Elderly patients have the same probability of surgical success as those who are younger. Motor or sensory decits of slight or moderate severity usually recover completely after surgery. Those which are severe may recover only partially or remain unchanged; the probability of recovery is inversely proportional to the severity and duration of impairment. In those which have lasted longer than one month, the longer the time which has elapsed from their onset, the lower the chances of complete recovery. Approximately 60% of the patients with a cauda equina syndrome who present with paralysis of the bladder obtain incomplete recovery of function. Patients with partial dysfunction nearly always achieve a good functional recovery. The interval between the onset of the syndrome and surgical treatment does not affect the outcome. Nevertheless, in the presence of a cauda equina syndrome, it is advisable to carry out the operation rapidly to avoid the risk that partial sphincter, motor or sensory dysfunction may worsen, thus decreasing the chances of a complete postoperative recovery. Most patients return to their preoperative work or to a lighter job, usually by eight to ten weeks after surgery, but earlier for sedentary rather than manual workers. Three factors have a signicant effect on the result of surgery, namely the preoperative duration of the clinical syndrome, the surgical ndings, and the degree of tension on the nerve root. Patients with pain for more than six months before surgery and those with annular bulging or
THE JOURNAL OF BONE AND JOINT SURGERY

MANAGEMENT OF HERNIATION OF THE LUMBAR DISC

573

Fig. 4a

Fig. 4b

Fig. 4c

Extravertebral approach. Figure 4a Exposure of the transverse processes and the intertransverse muscles. Figure 4b Excision of the intertransverse muscle and ligament. Figure 4c After retraction of the nerve root, the disc and the extruded or migrated disc fragment are seen.
60

contained disc herniation tend to obtain less satisfactory results than those with a shorter duration of pain and/or a migrated disc herniation. A satisfactory result is associated more often with markedly positive nerve-root tension tests. A clear-cut prevalence of low back pain over leg pain implies a high risk of an unsatisfactory outcome.

Lateral herniations
Lateral intraforaminal or extraforaminal herniations can be excised through the usual interlaminar technique or through approaches which lead to the outer aspect of the intervertebral foramen without passing through the vertebral 57 58 canal (Fig. 3). Para-articular, paraspinal and paral59 ateral extravertebral approaches have also been described (Fig. 4). Excision of an extruded intraforaminal herniation or a fragment of disc which has migrated into the root canal may be a demanding procedure under normal vision, particularly if the fragment is small or attempts are made to preserve much of the articular processes. Complete arthrectomy, however, which makes the removal of the herniation easier, may decrease the stability of the involved motion segment. These drawbacks are avoided when using the operating microscope (Fig. 3). All extravertebral approaches can be carried out under normal vision, but it may be difcult to identify the anatomical structures in the intervertebral foramen and handle the nerve root without trauma in order to visualise and excise the herniated disc. It is thus preferable to use the operating microscope or at least a frontal source of light. Indications for the interlaminar or extravertebral approaches. There appear to be no signicant differences in terms of clinical outcome between the interlaminar and
VOL. 81-B, NO. 4, JULY 1999

extravertebral approaches. With the former, there is a risk of compromising vertebral stability if too extensive or complete arthrectomy is carried out. This rarely occurs when using the operating microscope which also allows easier detection of a fragment of the disc which has migrated into the foramen. With extravertebral approaches, there is less risk of affecting the stability of the motion segment, but a greater chance of leaving behind a free fragment which may extend a considerable way into the vertebral canal. Extravertebral approaches do not allow for elimination of a concomitant stenosis of the root canal which may be seen in patients with a lateral herniation. With combined approaches the operation time is longer and the surgery more destructive; this is 61 associated with a higher rate of unsatisfactory results. Extravertebral approaches are indicated in contained or extruded extraforaminal herniations and in the rare cases in which a fragment of the disc has migrated largely or entirely outside the intervertebral foramen. In most intraforaminal herniations microdiscectomy is indicated through an interlaminar approach, but in these circumstances, a precise preoperative diagnosis is important to determine whether the herniation is contained, extruded or migrated, the location and size of a migrated fragment and whether stenosis is present. When this is the case, the interlaminar approach should always be preferred.

Recurrent herniation
Denition of terms. The term recurrent indicates a disc herniation occurring at the same level and side as the primary protrusion and new herniation describes a contralateral lesion at the same level or a fresh herniation at a different site. A herniation may be considered as a recur-

574

F. POSTACCHINI

Fig. 5a

Fig. 5b

Recurrent L4/L5 disc herniation on the left. Figure 5a Spin-echo T1-weighted axial image showing pathological tissue, isointense to the disc, anterior to the thecal sac. Figure 5b Spin-echo T1-weighted axial image obtained after injection of gadolinium. The pathological tissue shows signal enhancement in the peripheral, but not the central, portion (arrowhead). These ndings indicate a recurrent disc herniation surrounded by scar tissue.

rence if the radicular pain has entirely receded or considerably decreased for at least a few months after primary discectomy. If this is not the case it should be regarded as a failure of surgery. The time interval for a disc herniation to be considered as a recurrence may arbitrarily be esti27 mated at three months . Before then it is unlikely that the newly-formed tissue within the disc is the source of a prolapse causing nerve-root compression. Early recurrent herniations are those which occur between the third and the twelfth months after operation and late recurrences are those developing thereafter. Conservative treatment. The natural history of recurrent disc herniation is unknown and there is little information regarding the effectiveness of conservative management. 10 62 Hakelius and Naylor reported that 5% and 8%, respectively, of their patients who had undergone operation had complained of recurrent radicular pain which resolved with conservative treatments, but in both series the diagnosis was based on clinical evaluation. In my experience, patients with a small recurrent herniation tend to have a spontaneous resolution of lumboradicular symptoms, which may be hastened by conservative management for at least four months before further operation is considered. In these patients the result of further surgery is unpredictable. In patients with a medium or large recurrent herniation causing severe nerve-root compression, disappearance of the radicular pain is less likely to occur than in those with a primary herniation. This is due to the characteristics of the herniated tissue, which shows little tendency to decrease in size as a result of dehydration, and to the presence of epidural brosis which reduces the mobility of the affected nerve root. If lumboradicular symptoms persist after one to two months in these patients, conservative treatment should be withdrawn. This is particularly so in those who have obtained a satisfactory result after the primary discectomy followed by a long period free from pain.

Results of surgery. The proportion of satisfactory results after repeat discectomy is not certain, since in most of the studies, patients with a recurrence have been evaluated along with those presenting with a new herniation. In numerous series the proportion of satisfactory outcomes 63-68 69 ranged between 77% and 85%. Finnegan et al reported 92% of satisfactory results in patients with recurrent or 70 new herniation, but in another study, only 64% of patients who had a further operation at the same level achieved a satisfactory result. In the past it was difcult to differentiate between epidural brosis and recurrent herniation but now the latter can usually be diagnosed easily by MRI with gadolinium (Fig. 5). Two factors seem to play a role in predicting the surgical outcome, namely, the quality of the result after primary discectomy, and the preoperative diagnosis and/or the intraoperative findings. Patients with satisfactory results after primary discectomy have a high chance of obtaining a good 64,70 but the quality of the result outcome after reoperation, tends to be slightly worse after this latter procedure due to the persistence of low back pain and/or radicular symptoms. Small recurrent herniations and epidural brosis are more often associated with unsatisfactory results than medium-sized or large herniation. Unsatisfactory results are more likely to occur when the time interval between primary discectomy and reoperation is less than one year. References
1. Weber H, Holme I, Amlie E. The natural course of acute sciatica with nerve root symptoms in a double-blind placebo-controlled trial evaluating the effect of Piroxicam. Spine 1993;18:1433-8. 2. Fraser RD. Chymopapain for the treatment of intervertebral disc herniation: a preliminary report of a double-blind study. Spine 1982; 7:608-12. 3. Fraser RD. Chymopapain for the treatment of intervertebral disc herniation: the nal report of a double-blind study. Spine 1984;9: 815-8. 4. Bush K, Cowen N, Katz D, Gishen P. The natural history of sciatica associated with disc pathology: a prospective study with clinical and independent radiologic follow-up. Spine 1992;17:1205-12.
THE JOURNAL OF BONE AND JOINT SURGERY

MANAGEMENT OF HERNIATION OF THE LUMBAR DISC

575

5. Delauche-Cavallier M-C, Budet C, Laredo J-D, et al. Lumbar disc herniation: computed tomography scan changes after conservative treatment of nerve root compression. Spine 1992;17:927-33. 6. Maigne J-Y, Rime B, Deligne B. Computed tomographic follow-up study of forty-eight cases of nonoperatively treated lumbar intervertebral disc herniation. Spine 1992;17:1071-4. 7. Saal JA, Saal JS, Herzog RJ. The natural history of lumbar intervertebral disc extrusions treated nonoperatively. Spine 1990;15:683-6. 8. Komori H, Shinomiya K, Nakai O, et al. The natural history of herniated nucleus pulposus with radiculopathy. Spine 1996;21:225-9. 9. Saal JA, Saal JS. Nonoperative treatment of herniated lumbar intervertebral disc with radiculopathy: an outcome study. Spine 1989;14: 431-7. 10. Hakelius A. Prognosis in sciatica: a clinical follow-up of surgical and non-surgical treatment. Acta Orthop Scand Suppl 1970:129. 11. Weber H. Lumbar disc herniation: a controlled prospective study with ten years of observation. Spine 1983;8:131-40. 12. Shvartzman L, Weingarten E, Sherry H, Levin S, Persaud A. Costeffectiveness analysis of extended conservative therapy versus surgical intervention in the management of herniated lumbar intervertebral disc. Spine 1992;17:176-82. 13. Dabezies EJ, Langford K, Morris J, Shields CB, Wilkinson HA. Safety and efcacy of chymopapain (discase) in the treatment of sciatica due to a herniated nucleus pulposus: results of a randomized, double-blind study. Spine 1988;13:561-5. 14. Javid MJ, Nordby EJ, Ford LT, et al. Safety and efcacy of chymopapain (Chymodiactin) in herniated nucleus pulposus with sciatica: results of a randomised double-blind study. JAMA 1983;249: 2489-94. 15. Dabezies EJ, Beck C, Shoji H. Chymopapain in perspective. Clin Orthop 1986;206:10-4. 16. Dubuc FL, Apfelbach H, Simmons JW, Javid M, Nordby EJ. Four brief reports on long-term results of intradiscal chymopapain. Clin Orthop 1986;206:42-4. 17. Flanagan N, Smith L. Clinical studies of chemonucleolysis patients with 10-20 year follow-up evaluation. Clin Orthop 1986;206:15-7. 18. Gogan WJ, Fraser RD. Chymopapain: a 10-year, double-blind study. Spine 1992;17:388-94. 19. Jabaay GA. Chemonucleolysis: 8-10 year follow-up evaluation. Clin Orthop 1986;206:24-31. 20. Maciunas RJ, Onofrio BM. The long-term results of chymopapain chemonucleolysis for lumbar disc disease: ten-year follow-up results in 268 patients injected at the Mayo Clinic. J Neurosurg 1986;65:1-8. 21. Manseld F, Polivy K, Boyd R, Huddleston J. Long-term results of chymopapain injections. Clin Orthop 1986;206:67-9. 22. Nordby EJ. 8-13 year follow-up evaluation of chemonucleolysis patients. Clin Orthop 1986;206:18-23. 23. Postacchini F, Perugia D. Long term results of chemonucleolysis. J Bone Joint Surg [Br] 1992;74-B:Suppl 1,69. 24. Sutton JC Jr. Repeat chemonucleolysis. Clin Orthop 1986;206:45-9. 25. Thomas JC Jr, Wiltse LL, Widell EH Jr, et al. Chemonucleolysis: a ten-year retrospective study. Clin Orthop 1986;206:61-6. 26. Wilson LF, Mulholland RC. Results of chemonucleolysis: the longterm results of chemonucleolysis. In: Wittenberg RH, ed. Chemonucleolysis and related intradiscal therapies. Stuttgart, New York: Georg Thieme Verlag, 1994:89-126. 27. Postacchini F. Lumbar disc herniation. Wien, New York: Springer Wien, 1999. 28. Bocchi L, Ferrata P, Passarello F. The Onik method of automated percutaneous lumbar discectomy (APLD): criteria of selection, technique and evolution of results. Giorn Ital Ortop Traumat 1991;17: 5-22. 29. Bonaldi G, Belloni G, Prosetti D, Moschini L. Percutaneous discectomy using Oniks method: 3 years experience. Neuroradiology 1991;33:516-9. 30. Davis GW, Onik G. Clinical experience with automated percutaneous lumbar discectomy. Clin Orthop 1989;238:98-103. 31. Gill K, Blumenthal SL. Automated percutaneous discectomy: longterm clinical experience with the nucleotome system. Acta Orthop Scand Suppl 1993;251:30-3. 32. Onik G, Mooney V, Maroon JC, et al. Automated percutaneous discectomy: a prospective multi-institutional study. Neurosurgery 1990; 26:228-33.
VOL. 81-B, NO. 4, JULY 1999

33. Dullerud R, Amundsen T, Lie H, et al. Clinical results after percutaneous automated lumbar nucleotomy: a follow-up study. Acta Radiol 1995;36:418-24. 34. Grevitt MP, McLaren A, Shackleford IM, Mulholland RCM. Automated percutaneous lumbar discectomy: an outcome study. J Bone Joint Surg [Br] 1995;77-B:626-9. 35. Kahanovitz N, Viola K, Goldstein T, Dawson E. A multicenter analysis of percutaneous discectomy. Spine 1990;15:713-5. 36. Revel M, Payan C, Vallee C, et al. Automated percutaneous lumbar discectomy versus chemonucleolysis in the treatment of sciatica: a randomised multicenter trial. Spine 1993;18:1-7. 37. Shapiro S. Long-term follow up of 57 patients undergoing automated percutaneous discectomy. J Neurosurg 1995;83:31-3. 38. Dullerud R, Amundsen T, Nakstad PH, Magnaes B. CT changes after lumbar percutaneous automated nucleotomy. Acta Radiol 1994; 35:409-14. 39. Hijikata S. Percutaneous nucleotomy: a new concept technique and 12 years experience. Clin Orthop 1989;238:9-23. 40. Hoppenfeld S. Percutaneous removal of herniated lumbar discs: 50 cases with ten-year follow-up periods. Clin Orthop 1989;238:92-7. 41. Kambin P. Arthroscopic microdiscectomy. Arthroscopy 1992;8: 287-95. 42. Mayer HM, Brock M. Percutaneous endoscopic discectomy: surgical technique and preliminary results compared to microsurgical discectomy. J Neurosurg 1993;78:216-25. 43. Ohnmeiss DD, Guyer RD, Hochschuler SH. Laser disc decompression: the importance of proper patient selection. Spine 1994;18: 2054-9. 44. Gleave JR, Macfarlane R. Prognosis for recovery of bladder function following lumbar central disc prolapse. Br J Neurosurg 1990;4: 205-9. 45. Kostuik JP, Harrington I, Alexander D, et al. Cauda equina syndrome and lumbar disc herniation. J Bone Joint Surg [Am] 1986; 68-A:386-91. 46. Spannare BJ. Prolapsed lumbar intervertebral disc with partial or total occlusion of the spinal canal: a study of 30 patients with and 28 without cauda equina symptoms. Acta Neurochir Wien 1978;42: 189-98. 47. Caspar W, Campbell B, Barbier DD, Kretschmmer R, Goffried Y. The Caspar microsurgical discectomy and comparison with a conventional standard lumbar disc procedure. Neurosurgery 1991;28:78-87. 48. Loew F, Caspar W. Surgical approach to the lumbar disc herniations. Adv Tech Stand Neurosurg 1978;5:153-74. 49. McCulloch JA. Principles of microsurgery for lumbar disc disease. New York: Raven Press, 1989. 50. Ebeling U, Reichenbach W, Reulen H-J. Results of microsurgical lumbar discectomy: review on 485 cases. Acta Neurochir Wien 1986; 81:45-52. 51. Goald HJ. Microlumbar discectomy: follow-up of 477 patients. J Microsurg 1980;2:95-100. 52. Postacchini F, Cinotti G, Perugia D. Microdiscectomy in treatment of herniated lumbar disc. Ital J Orthop Traumatol 1992;18:5-16. 53. Sivers HR, Lewis PJ, Asch HL, Clabeaux DE. Lumbar discectomy for recurrent disk herniation. J Spinal Disord 1994;7:408-19. 54. Thomas AMC, Afshar F. The microsurgical treatment of lumbar disc protrusion. J Bone Joint Surg [Br] 1987;69-B:696-8. 55. Williams RW. Microlumbar discectomy: a 12-year statistical review. Spine 1986;11:851-2. 56. Zahrawi F. Microlumbar discectomy: is it safe as an outpatient procedure? Spine 1994;19:1070-4. 57. Reulen HJ, Pfaundler S, Ebeling U. The lateral microsurgical approach to the extracanalicular lumbar disc herniation. I: a technical note. Acta Neurochir Wien 1987;84:64-7. 58. Recoules-Arches D. La hernie discale lombaire du canal de conjugaison: propos de 37 hernies op r es. Neurochirurgie 1984;30: ee 301-7. 59. Ray CD. The paralateral approach to lumbar decompression. Am Assoc Nurol Surg Annu Meet 1983;191(poster session, abstract 21 W). 60. Postacchini F, Cinotti G, Gumina S. Microsurgical excision of lateral lumbar disc herniation through an interlaminar approach. J Bone Joint Surg [Br] 1998;80-B:201-7.

576

F. POSTACCHINI

61. Faust SE, Ducker TB, Van Hassent JA. Lateral lumbar disc herniations. J Spinal Disord 1992;5:97-103. 62. Naylor A. The late results of laminectomy for lumbar disc prolapse: a review after ten to twenty-ve years. J Bone Joint Surg [Br] 1974; 56-B:17-29. 63. Cinotti G, Roysam GS, Eisenstein SM, Postacchini F. Ipsilateral recurrent lumbar disc herniation: a prospective, controlled study. J Bone Joint Surg [Br] 1998;80-B:825-32. 64. Ebeling U, Kalbaryck H, Reulen HJ. Microsurgical reoperation following lumbar disc surgery: timing, surgical ndings, and outcome in 92 patients. J Neurosurg 1989;70:397-404. 65. Epstein JA, Lavine LS, Epstein BS. Recurrent herniation of the lumbar intervertebral disk. Clin Orthop 1967;52:169-78.

66. J nsson B, Str mqvist B. Repeat decompression of lumbar nerve o o roots: a prospective two-year evaluation. J Bone Joint Surg [Br] 1993; 75-B:894-7. 67. Kim SS, Michelsen CB. Revision surgery for failed back surgery syndrome. Spine 1992;17:957-60. 68. Martin G. Recurrent disc prolapse as a cause of recurrent pain after laminectomy for lumbar disc lesions. NZ Med J 1980;91:206-8. 69. Finnegan WJ, Fenlin JM, Marvel JP, Nardini RJ, Rothman RH. Results of surgical intervention in the symptomatic multiply-operated back patient: analysis of sixty-seven cases followed for three to seven years. J Bone Joint Surg [Am] 1979;61-A:1077-82. 70. OSullivan MG, Connolly AE, Buckley TF. Recurrent lumbar disc protrusion. Br J Neurosurg 1990;4:319-25.

THE JOURNAL OF BONE AND JOINT SURGERY