Archives of Orthopaedic and Traumatic Surgery
Arch Orthop Traumat Surg 93, 103-110 (1979)
© J F Bergmann Verlag 1979
Lumbosacral Fusion by Standing H-Graft Technique Turkka Tunturi', Timo Paakkala, Hannu Ptiala', Tapio Tervo 2, and Pentti Rokkanen' ' Institute of Clinical Sciences, University and Central Hospital, Tampere, Teiskontie 35, SF-33520 Tampere 52 2 Central Hospital of Middle Finland, Jyvaskyla, Finland
Summary Fusion of the lumbosacral spine was carried out on 44 patients A modification of the standing Hgraft technique was used in 41 of the patients The operative diagnosis was degeneration of the lumbosacral spine in nearly two-thirds of the cases and spondylolisthesis in about one-third Operations for low back or sciatic pain had previously been performed on 17 patients The duration of postoperative bed rest ranged from one to six weeks; a lumbosacral corset was used at least for six weeks postoperatively. The average stay in hospital was 52 days Complications occurred in five cases. The operative results of 40 patients were evaluated on the basis of a follow-up examination performed on average 3 6 years after operation Nearly half the patients with radiologically successful fusions had a good subjective improvement and about one-fourth had returned to their previous or corresponding occupations The operative result was at least fair in about half the patients, assessed by a measure of the patients' subjective improvement and working capacity In the non-fusion group (six cases) the operative results were exclusively poor Age over 40 years, long-standing preoperative disability and previous back operations proved prognostic factors for poor results. Assessed from functional radiographs, 63 % of two-segment fusions and 95 % of one-segment fusions were successful. The operative technique appeared simple and suitable for fusions of one spinal segment, but the value of the lumbosacral fusion in the treatment of low back pain seemed doubtful because of the great proportion of poor clinical results even following technically successful fusions. Zusammenfassung Die dorsale Spondylodese der Lumbosakralregion wurde an 44 Patienten durchgefuihrt.
Eine Modifizierung der stehenden H-Span-Technik wurde bei 41 Patienten angewendet Lumbosakrale Degeneration war in fast zwei Drittel, Spondylolisthesis in etwa einem Drittel der Flle die Operationsdiagnose Bei 17 Patienten waren bereits friiher Operationen wegen Riickenschmerzen oder Ischiassyndrom durchgefiihrt worden Die postoperative Bettruhe dauerte eine bis sechs Wochen Nach der Operation wurde ein lumbosakrales Korsett mindestens sechs Wochen getragen Der durchschnittliche Krankenhausaufenthalt betrug 52 Tage In fnf Fillen kamen Komplikationen vor. Die Operationsergebnisse von 40 Patienten wurden auf Grund der Nachuntersuchung bewertet Die Beobachtungszeit betrug im Durchschnitt 3,6 Jahre Beinahe die Halfte der Patienten mit rntgenologisch stabiler Fusion hatte ein gutes subjektives Endergebnis, und ein Viertel hatten ihre friihere oder eine entsprechende Arbeit wieder aufgenommen Bei der Anwendung der subjektiven Symptome und der Arbeitsfahigkeit als Kriterien der Bewertung war das operative Resultat wenigstens ,,mf13ig" bei etwa der Halfte der Patienten Bei den sechs Fillen, bei denen die stabile Fusion ausblieb, waren die Operationsresultate ausschlie13lich schlecht Alter fiber 40 Jahre, langdauernde voroperative Arbeitsunfahigkeit und friihere Riickenoperationen beeintrachtigten die Operationserfolge. Auf Grund der Bewertung funktioneller Rontgenaufnahmen waren 63 % der Versteifungen zweier Segmente und 95 % der Versteifungen eines Segmentes erfolgreich. Die angewandte Operationstechnik erwies sich als einfach und angebracht fiir die Versteifung eines, aber nicht mehrerer Segmente In der Behandlung von Kreuzschmerzen scheint aber der Wert der lumbosakralen Spondylodese fraglich zu sein, weil die klinischen Endergebnisse, selbst bei technisch erfolgreichen Fusionen, zum gr 13 Bten Teil schlecht waren.
Offprint requests to: T Tunturi, M D (address see above)
0344-8444/79/0093/0103/$ 1 60
T Tunturi et al : Lumbosacral Fusion by Standing H-Graft Technique
104
Various techniques of lumbosacral fusion have been used in the treatment of chronic low back pain due to instability caused by degenerative diseases and anomalies of the lumbosacral spine It would be ideal to perform fusion by a simple technique with a high frequency of success, a short postoperative stay in hospital and a minimum of complications According to the literature, fusion by the standing H-graft technique (Besa, 1960 ; Langenskibld, 1967) may fulfill these criteria, but there is not much reported data available concerning this method. The present paper is an analysis of the operative results obtained by a modification of the standing Hgraft technique: the complications, the fusion rate and the follow-up of the patients with regard to working capacity and subjective improvement The results of a statistical analysis are presented to delineate factors influencing the operative results.
Patients and Methods Patients The series consisted of 44 patients subjected to fusion of the lumbosacral spine, mainly by the standing H-graft technique (see below), at the Central Hospital of Middle Finland between 1972-1975. At the time of operation the mean age of the patients was 37 years, ranging from 20 to 47 years; 17 patients were female, 27 male Manual labour was represented preoperatively by 38 patients, sedentary work by six Heavy manual labour was performed by 27 patients, moderately heavy by six and light labour by five. All the patients had suffered from low back pain causing disability or marked restriction in daily activities The average
duration of preoperative symptoms was 9 8 years One patient had experienced low back pain for less than one year, over one-third for at least 10 years One-third of the patients had been disabled by the back pain for more than one year before operation. All the patients had been conservatively treated Rest was the most frequently used and the most effective therapy; it had brought relief in two-thirds of the cases. A total of 22 back operations had previously been performed on 17 patients; hemilaminectomy on one, lumbosacral fusion and hemilaminectomy on one, one disc operation on 12, two disc operations on two patients and four disc operations on one.
The operative diagnosis was degeneration of the lumbosacral spine in nearly two-thirds of the patients, spondylolysis or spondylolisthesis in about one-third (Table 1).
Standing H-Graft Technique The operation was performed under general anesthesia with the patient prone A straight skin incision was made in the midline from L4 to 52 The fascia was split from both sides of the spinous processes The muscles were drawn aside The supraspinous ligament was cut at the level of the lower margin of the spinous process of the lowest vertebra to be fused and detached as far as the upper margin of the spinous process of the uppermost vertebra to be fused The interspinous ligament was removed The laminae and the spinous processes were freshened The spinous processes were thinned and shortened. From the posterior iliac crest a bone graft of suitable size was taken and immediately shaped by removing the cancellous bone between the cortical laminae at both ends (Fig 1A) The spinal lordosis was straightened by adjustment of the operation table The posture was slowly restored, while placing the bone graft so that it became wedged between the spinous processes (Fig I B). Cancellous bone from the iliac crest was applied on and between the laminae If excision or exploration of a disc herniation was performed, cancellous bone was applied to the other
A Table 1 Operative diagnoses of 44 patients Diagnosis
Number of patients
Spondylolysis and spondylolisthesisa Grade I Grade II Grade III-IV
15 b
Degeneration of the lumbosacral spine Disc degeneration Sequel of previous disc excision
27
8 4 0
12 ' 15 d
Lumbarization-sacralization
I
Unsuccessful previous fusion
1
Total a
B
44
Meyerding's (1932) classification Radiographs of three patients not available Disc herniation in one case Recurrent disc herniation in two cases
Fig 1 A The principle of removing cancellous bone between the cortical laminae B H-graft wedged between two spinous processes
105
T Tunturi et al : Lumbosacral Fusion by Standing H-Graft Technique Table 2 Fused segments in 44 patients
Table 3 Criteria of evaluation of operative results
Fused segment (s)
Criteria of evaluation
L 3-L5
Number of patients 1
L 4-L5
15
L 3 51 L 4-S 1
1 19 a
L 5 51 Total a
Change of subjective symptoms Much better Better No change Worse
8b 44
5 3 1 O
Symptoms at follow-up (only the lowest score is considered)
No pain Pain in heavy work Pain after lengthy sitting Pain in light work Persistent pain
Two cases, not H-graft technique bOne case, not H-graft technique
side only The supraspinous ligament was sutured in position and the wound was closed. Fusion was performed by this technique in 41 patients and by a combination (Ryerson, 1932) of the methods described by Hibbs (1911) and Albee (1911) in three cases, not suitable for the H-graft technique Fusion was performed on one spinal segment in 23 cases and two segments in 20 cases (Table 2) In addition to fusion, exploration of an intervertebral space was performed in 13 cases and excision of a herniated disc in three. The average duration of the operation was 1 h and 47 min with a range from 1 to 3 h The average loss of blood was about 600 g, ranging from 150 to 2500 g Per and postoperatively 18 patients received blood, on average about 1100 g.
Scores
5 3 2 1 0
Working capacity Previous or a corresponding work Lighter work owing to the back pain Lighter work for other reasons
5 2
Pensioned because of the back pain Pensioned for other reasons
O
a
Succeeds in work (pensioner in daily activities) Well Fairly well Poorly
5 3 O
Same score as for 'symptoms at follow-up'
a
Score for the patient-Operative Result
Postoperative Treatment The duration of bed rest after one-segment fusion was one week in eight cases, three weeks in eight, and six weeks in seven After fusion of more than one spinal segment all patients were nursed in bed for six weeks No prophylaxis for venous thrombosis was used The average stay in hospital was 52 days, ranging from 16 to 109 days. Postoperatively the patients visited the outpatient department an average of seven times, ranging from 2 to 29 times. Postoperatively 39 patients used a lumbosacral corset, 8 for 6 weeks, 11 for 12 weeks, and 20 for a longer period.
Complications During the stay in hospital complications developed in five cases: one patient died of pulmonary embolism on the seventh postoperative day, deep venous thrombosis of the lower extremity was diagnosed in one patient, and wound infection in three. No late complications occurred.
Follow-Up Examination To the data obtained from the patient's records was added the information collected at follow-up, including the results of clinical and radiological examinations Forty patients (91%) attended for follow-up examination Four patients did not attend: two for unknown reasons, one had died of pulmonary embolism postoperatively, and one had committed suicide six months after operation.
16-20
Good
9-15
Fair
0-8
Poor
The average interval between operation and follow-up examination was 3 6 years, ranging from one year and seven months to five years and three months.
Evaluation of the Operative Results The operative results were assessed by scores given on the criteria presented in Table 3 (Tunturi et al , in press) The technical success of fusion was not considered in evaluating the operative result.
Radiological Examination The patients attending follow-up were subjected to functional radiography of the lumbosacral spine One antero-posterior radiograph and two lateral radiographs were taken with the patient sitting, one in extension, the other in flexion Preoperative or immediately postoperative radiographs were available in 39 cases for comparison. Intervertebral movement in the fused segment(s) was regarded as a sign of non-fusion A change by at least 2 mm in height of the anterior portion of the intervertebral space or a shift by at least 2 mm in the horizontal plane in the functional radiographs were considered the criteria of movement Furthermore, a change in the gap between the H-graft and the
106
T Tunturi et al : Lumbosacral Fusion by Standing H-Graft Technique
spinous processes in the functional radiographs was considered as evidence of non-fusion. The degree of spondylolisthesis was classified according to Meyerding (1932) (Table 1).
Statistical Methods In testing statistical hypotheses the chi-square and t-test were used.
Table 4 Operative results compared with age at operation in 30 patients with radiologically successful fusions Age (years)
Number Operative result of patients Good Fair Poor
21-30
1
31-40 41-50
15 14
3 2
8 2
4 10
Total
30
5
11
14
Mean score
1
15
11.1 I 11 3 11 1 70 93
Results ClinicalResults The clinical results were analyzed in two groups depending on the radiological evidence of successful fusion Thirty patients had the standing H-graft technique and had a radiologically successful fusion. Subjective Improvement At follow-up, 14 patients reported that their back was much better than before operation, six that it was somewhat better Six patients' symptoms were unchanged and four patients stated that their back was worse than before operation. The back was painless in one case, became painful on heavy work in four, after lengthy sitting in 11, and on light work in six cases Eight of the patients stated that they had persistent back pain. The fused area was painful in 18 patients, the area above it in three and the area below it in one patient. Pain radiating to the thighs was experienced by 18 patients. Working Capacity The duration of postoperative disability was less than five months in three cases, from five to seven months in three, from seven to 12 months in six and over 12 months in 18 cases. Eight patients had resumed their previous or a corresponding occupation Five patients had shifted to lighter work because of back pain Sixteen patients were disabled owing to back pain and one for another reason at the time of follow-up. Of those 17 patients who were pensioned on account disability, 11 managed at least fairly well in their daily activities Of those 8 patients who were working in their previous or a corresponding occupation, all succeeced at least fairly well. Seven out of 16 patients under 40 years old and nine out of 13 patients over 40 were disabled owing to back pain. The previous or a corresponding work had been resumed by one out of 16 laborers with heavy manual occupations and by seven out of 13 patients with other occupations. Three out of 16 patients with previous back operations worked in their former or a corresponding occu-
Table 5 Operative results compared with kind of preoperative work in 30 patients with radiologically successful fusions Kind of work
Heavy manual labourlabour Moderately heavy manual labour Light manual labour Sedentary work Total
Number Operative result of patients Good Fair Poor
Mean score
16
1
7
8
84
6
1
3
2
97
5
2
1
2
11 4
3
1
2
97
30
5
14
93
11
pation, one had shifted to lighter work and 11 were disabled owing to back pain and one for another reason. Operative Results and Factors Influencing Them Assessed by the criteria adopted (Table 3), the result was good in five cases ( 17%), fair in 11 (37 %) and poor in 14 ( 47%). Among the patients under 40 years old, the result was at least fair in 75 % and poor in 25 % The corresponding figures for those over 40 were 29 % and 71 % (Table 4) The patients over 40 had statistically significantly (P< 0 025) poorer results than the patients under 40. A poor result was noticed in half the labourers in heavy work and in about one-third of those performing moderately heavy or light manual work preoperatively (Table 5) The difference was not statistically significant because of the deviation of the scores. The duration of symptoms was over 10 years in 14 cases The result was good in two of these, fair in four, and poor in eight The corresponding figures for 16 patients with back pain for less than 10 years were three, seven, and six The difference in operative results between these groups was not statistically significant.
T Tunturi et al : Lumbosacral Fusion by Standing H-Graft Technique Table 6 Operative results compared with preoperative duration of disability in 30 patients with radiologically successful fusions Preoperative disability dions (months)
Number Operative result of patients Good Fair Good Fair Poor
Mean score
1-6 6-12 > 12
6 13 11
2 3
4 5 2
5 9
15 2 10 6 45
Total
30
5
11
14
93
Table 7 Operative results compared with the number of previous back operations in 30 patients with radiologically successful fusions Number of Number Operative result previous back of patients Good Fair operations Good Fair Poor operations
Mean score
0 _1
14 16
2 3
8 3
4 10
11 4 74
Total
30
5
11
14
93
Table 8 Operative results compared with diagnoses in 30 patients with radiologically successful fusions Operative diagnosis Spondylolysis and spondylolisthesis Disc degeneration Sequel of previous disc excision Lumbarization-
Number Operative result of patients Good Fair Poor
Mean score
Of those 16 patients who had undergone previous back operations, nearly two-thirds had a poor result (Table 7) In this group the operative results were statistically significantly (P< 0 05) poorer than in the patients with no previous back surgery. The proportion of poor results was lower in the group of patients operated on for spondylolysis and spondylolisthesis than in the other groups (Table 8). In the group of one-segment fusions (20 patients) a good result was noted in four patients, a fair result in six, and a poor result in ten Of those ten patients who had undergone multiple level fusion, one had a good result, five a fair result, and four a poor result. Radiologically Unsuccessful Fusions Of those seven patients in whom fusion was radiologically unsuccessful, one showed non-fusion in the segment L 5-sacrum, to which the fusion was extended, although the segment under treatment was L 4-L 5 This patient had a good result of operation In the remaining six patients the result was poor, assessed on the principles adopted (Table 3) Two of these patients reported that their back had become somewhat better, while the symptoms were unchanged in three and back pain was aggravated in one One patient had shifted to lighter work and five were disabled owing to back pain The mean score for the patients showing non-fusion was 3.8 This was statistically significantly (P< 0 0025) lower than the score for the group of patients with successful fusions.
5
1
3
1
12 8
RadiologicalResults
10 13
1 3
5 2
4 8
10 6 73
30
5
11
14
93
Out of 40 patients attending the follow-up examination, 37 had been operated on by the standing H-graft technique Fusion was radiologically successful in 30 of these patients, while non-fusion was observed in seven. One-segment fusion was successful in 20 ( 95 %) out of 21 patients (in 94 % of patients immobilized for 1-3 weeks, in 100% of patients immobilized for six weeks). Multiple level fusion was successful in 10 (63 %) out of 16 patients One-segment fusions were statistically significantly (P< 0 025) more often successful than multiple level fusions.
sacralization
Unsuccessful previous fusion
Total
107
The operative results are compared with preoperative duration of disability in Table 6 The differences in results between all three groups were statistically significant (between first and second P < 0 005, second and third P< 0 0025, first and third P
Arch Orthop Traumat Surg 93, 103-110 (1979)
© J F Bergmann Verlag 1979
Lumbosacral Fusion by Standing H-Graft Technique Turkka Tunturi', Timo Paakkala, Hannu Ptiala', Tapio Tervo 2, and Pentti Rokkanen' ' Institute of Clinical Sciences, University and Central Hospital, Tampere, Teiskontie 35, SF-33520 Tampere 52 2 Central Hospital of Middle Finland, Jyvaskyla, Finland
Summary Fusion of the lumbosacral spine was carried out on 44 patients A modification of the standing Hgraft technique was used in 41 of the patients The operative diagnosis was degeneration of the lumbosacral spine in nearly two-thirds of the cases and spondylolisthesis in about one-third Operations for low back or sciatic pain had previously been performed on 17 patients The duration of postoperative bed rest ranged from one to six weeks; a lumbosacral corset was used at least for six weeks postoperatively. The average stay in hospital was 52 days Complications occurred in five cases. The operative results of 40 patients were evaluated on the basis of a follow-up examination performed on average 3 6 years after operation Nearly half the patients with radiologically successful fusions had a good subjective improvement and about one-fourth had returned to their previous or corresponding occupations The operative result was at least fair in about half the patients, assessed by a measure of the patients' subjective improvement and working capacity In the non-fusion group (six cases) the operative results were exclusively poor Age over 40 years, long-standing preoperative disability and previous back operations proved prognostic factors for poor results. Assessed from functional radiographs, 63 % of two-segment fusions and 95 % of one-segment fusions were successful. The operative technique appeared simple and suitable for fusions of one spinal segment, but the value of the lumbosacral fusion in the treatment of low back pain seemed doubtful because of the great proportion of poor clinical results even following technically successful fusions. Zusammenfassung Die dorsale Spondylodese der Lumbosakralregion wurde an 44 Patienten durchgefuihrt.
Eine Modifizierung der stehenden H-Span-Technik wurde bei 41 Patienten angewendet Lumbosakrale Degeneration war in fast zwei Drittel, Spondylolisthesis in etwa einem Drittel der Flle die Operationsdiagnose Bei 17 Patienten waren bereits friiher Operationen wegen Riickenschmerzen oder Ischiassyndrom durchgefiihrt worden Die postoperative Bettruhe dauerte eine bis sechs Wochen Nach der Operation wurde ein lumbosakrales Korsett mindestens sechs Wochen getragen Der durchschnittliche Krankenhausaufenthalt betrug 52 Tage In fnf Fillen kamen Komplikationen vor. Die Operationsergebnisse von 40 Patienten wurden auf Grund der Nachuntersuchung bewertet Die Beobachtungszeit betrug im Durchschnitt 3,6 Jahre Beinahe die Halfte der Patienten mit rntgenologisch stabiler Fusion hatte ein gutes subjektives Endergebnis, und ein Viertel hatten ihre friihere oder eine entsprechende Arbeit wieder aufgenommen Bei der Anwendung der subjektiven Symptome und der Arbeitsfahigkeit als Kriterien der Bewertung war das operative Resultat wenigstens ,,mf13ig" bei etwa der Halfte der Patienten Bei den sechs Fillen, bei denen die stabile Fusion ausblieb, waren die Operationsresultate ausschlie13lich schlecht Alter fiber 40 Jahre, langdauernde voroperative Arbeitsunfahigkeit und friihere Riickenoperationen beeintrachtigten die Operationserfolge. Auf Grund der Bewertung funktioneller Rontgenaufnahmen waren 63 % der Versteifungen zweier Segmente und 95 % der Versteifungen eines Segmentes erfolgreich. Die angewandte Operationstechnik erwies sich als einfach und angebracht fiir die Versteifung eines, aber nicht mehrerer Segmente In der Behandlung von Kreuzschmerzen scheint aber der Wert der lumbosakralen Spondylodese fraglich zu sein, weil die klinischen Endergebnisse, selbst bei technisch erfolgreichen Fusionen, zum gr 13 Bten Teil schlecht waren.
Offprint requests to: T Tunturi, M D (address see above)
0344-8444/79/0093/0103/$ 1 60
T Tunturi et al : Lumbosacral Fusion by Standing H-Graft Technique
104
Various techniques of lumbosacral fusion have been used in the treatment of chronic low back pain due to instability caused by degenerative diseases and anomalies of the lumbosacral spine It would be ideal to perform fusion by a simple technique with a high frequency of success, a short postoperative stay in hospital and a minimum of complications According to the literature, fusion by the standing H-graft technique (Besa, 1960 ; Langenskibld, 1967) may fulfill these criteria, but there is not much reported data available concerning this method. The present paper is an analysis of the operative results obtained by a modification of the standing Hgraft technique: the complications, the fusion rate and the follow-up of the patients with regard to working capacity and subjective improvement The results of a statistical analysis are presented to delineate factors influencing the operative results.
Patients and Methods Patients The series consisted of 44 patients subjected to fusion of the lumbosacral spine, mainly by the standing H-graft technique (see below), at the Central Hospital of Middle Finland between 1972-1975. At the time of operation the mean age of the patients was 37 years, ranging from 20 to 47 years; 17 patients were female, 27 male Manual labour was represented preoperatively by 38 patients, sedentary work by six Heavy manual labour was performed by 27 patients, moderately heavy by six and light labour by five. All the patients had suffered from low back pain causing disability or marked restriction in daily activities The average
duration of preoperative symptoms was 9 8 years One patient had experienced low back pain for less than one year, over one-third for at least 10 years One-third of the patients had been disabled by the back pain for more than one year before operation. All the patients had been conservatively treated Rest was the most frequently used and the most effective therapy; it had brought relief in two-thirds of the cases. A total of 22 back operations had previously been performed on 17 patients; hemilaminectomy on one, lumbosacral fusion and hemilaminectomy on one, one disc operation on 12, two disc operations on two patients and four disc operations on one.
The operative diagnosis was degeneration of the lumbosacral spine in nearly two-thirds of the patients, spondylolysis or spondylolisthesis in about one-third (Table 1).
Standing H-Graft Technique The operation was performed under general anesthesia with the patient prone A straight skin incision was made in the midline from L4 to 52 The fascia was split from both sides of the spinous processes The muscles were drawn aside The supraspinous ligament was cut at the level of the lower margin of the spinous process of the lowest vertebra to be fused and detached as far as the upper margin of the spinous process of the uppermost vertebra to be fused The interspinous ligament was removed The laminae and the spinous processes were freshened The spinous processes were thinned and shortened. From the posterior iliac crest a bone graft of suitable size was taken and immediately shaped by removing the cancellous bone between the cortical laminae at both ends (Fig 1A) The spinal lordosis was straightened by adjustment of the operation table The posture was slowly restored, while placing the bone graft so that it became wedged between the spinous processes (Fig I B). Cancellous bone from the iliac crest was applied on and between the laminae If excision or exploration of a disc herniation was performed, cancellous bone was applied to the other
A Table 1 Operative diagnoses of 44 patients Diagnosis
Number of patients
Spondylolysis and spondylolisthesisa Grade I Grade II Grade III-IV
15 b
Degeneration of the lumbosacral spine Disc degeneration Sequel of previous disc excision
27
8 4 0
12 ' 15 d
Lumbarization-sacralization
I
Unsuccessful previous fusion
1
Total a
B
44
Meyerding's (1932) classification Radiographs of three patients not available Disc herniation in one case Recurrent disc herniation in two cases
Fig 1 A The principle of removing cancellous bone between the cortical laminae B H-graft wedged between two spinous processes
105
T Tunturi et al : Lumbosacral Fusion by Standing H-Graft Technique Table 2 Fused segments in 44 patients
Table 3 Criteria of evaluation of operative results
Fused segment (s)
Criteria of evaluation
L 3-L5
Number of patients 1
L 4-L5
15
L 3 51 L 4-S 1
1 19 a
L 5 51 Total a
Change of subjective symptoms Much better Better No change Worse
8b 44
5 3 1 O
Symptoms at follow-up (only the lowest score is considered)
No pain Pain in heavy work Pain after lengthy sitting Pain in light work Persistent pain
Two cases, not H-graft technique bOne case, not H-graft technique
side only The supraspinous ligament was sutured in position and the wound was closed. Fusion was performed by this technique in 41 patients and by a combination (Ryerson, 1932) of the methods described by Hibbs (1911) and Albee (1911) in three cases, not suitable for the H-graft technique Fusion was performed on one spinal segment in 23 cases and two segments in 20 cases (Table 2) In addition to fusion, exploration of an intervertebral space was performed in 13 cases and excision of a herniated disc in three. The average duration of the operation was 1 h and 47 min with a range from 1 to 3 h The average loss of blood was about 600 g, ranging from 150 to 2500 g Per and postoperatively 18 patients received blood, on average about 1100 g.
Scores
5 3 2 1 0
Working capacity Previous or a corresponding work Lighter work owing to the back pain Lighter work for other reasons
5 2
Pensioned because of the back pain Pensioned for other reasons
O
a
Succeeds in work (pensioner in daily activities) Well Fairly well Poorly
5 3 O
Same score as for 'symptoms at follow-up'
a
Score for the patient-Operative Result
Postoperative Treatment The duration of bed rest after one-segment fusion was one week in eight cases, three weeks in eight, and six weeks in seven After fusion of more than one spinal segment all patients were nursed in bed for six weeks No prophylaxis for venous thrombosis was used The average stay in hospital was 52 days, ranging from 16 to 109 days. Postoperatively the patients visited the outpatient department an average of seven times, ranging from 2 to 29 times. Postoperatively 39 patients used a lumbosacral corset, 8 for 6 weeks, 11 for 12 weeks, and 20 for a longer period.
Complications During the stay in hospital complications developed in five cases: one patient died of pulmonary embolism on the seventh postoperative day, deep venous thrombosis of the lower extremity was diagnosed in one patient, and wound infection in three. No late complications occurred.
Follow-Up Examination To the data obtained from the patient's records was added the information collected at follow-up, including the results of clinical and radiological examinations Forty patients (91%) attended for follow-up examination Four patients did not attend: two for unknown reasons, one had died of pulmonary embolism postoperatively, and one had committed suicide six months after operation.
16-20
Good
9-15
Fair
0-8
Poor
The average interval between operation and follow-up examination was 3 6 years, ranging from one year and seven months to five years and three months.
Evaluation of the Operative Results The operative results were assessed by scores given on the criteria presented in Table 3 (Tunturi et al , in press) The technical success of fusion was not considered in evaluating the operative result.
Radiological Examination The patients attending follow-up were subjected to functional radiography of the lumbosacral spine One antero-posterior radiograph and two lateral radiographs were taken with the patient sitting, one in extension, the other in flexion Preoperative or immediately postoperative radiographs were available in 39 cases for comparison. Intervertebral movement in the fused segment(s) was regarded as a sign of non-fusion A change by at least 2 mm in height of the anterior portion of the intervertebral space or a shift by at least 2 mm in the horizontal plane in the functional radiographs were considered the criteria of movement Furthermore, a change in the gap between the H-graft and the
106
T Tunturi et al : Lumbosacral Fusion by Standing H-Graft Technique
spinous processes in the functional radiographs was considered as evidence of non-fusion. The degree of spondylolisthesis was classified according to Meyerding (1932) (Table 1).
Statistical Methods In testing statistical hypotheses the chi-square and t-test were used.
Table 4 Operative results compared with age at operation in 30 patients with radiologically successful fusions Age (years)
Number Operative result of patients Good Fair Poor
21-30
1
31-40 41-50
15 14
3 2
8 2
4 10
Total
30
5
11
14
Mean score
1
15
11.1 I 11 3 11 1 70 93
Results ClinicalResults The clinical results were analyzed in two groups depending on the radiological evidence of successful fusion Thirty patients had the standing H-graft technique and had a radiologically successful fusion. Subjective Improvement At follow-up, 14 patients reported that their back was much better than before operation, six that it was somewhat better Six patients' symptoms were unchanged and four patients stated that their back was worse than before operation. The back was painless in one case, became painful on heavy work in four, after lengthy sitting in 11, and on light work in six cases Eight of the patients stated that they had persistent back pain. The fused area was painful in 18 patients, the area above it in three and the area below it in one patient. Pain radiating to the thighs was experienced by 18 patients. Working Capacity The duration of postoperative disability was less than five months in three cases, from five to seven months in three, from seven to 12 months in six and over 12 months in 18 cases. Eight patients had resumed their previous or a corresponding occupation Five patients had shifted to lighter work because of back pain Sixteen patients were disabled owing to back pain and one for another reason at the time of follow-up. Of those 17 patients who were pensioned on account disability, 11 managed at least fairly well in their daily activities Of those 8 patients who were working in their previous or a corresponding occupation, all succeeced at least fairly well. Seven out of 16 patients under 40 years old and nine out of 13 patients over 40 were disabled owing to back pain. The previous or a corresponding work had been resumed by one out of 16 laborers with heavy manual occupations and by seven out of 13 patients with other occupations. Three out of 16 patients with previous back operations worked in their former or a corresponding occu-
Table 5 Operative results compared with kind of preoperative work in 30 patients with radiologically successful fusions Kind of work
Heavy manual labourlabour Moderately heavy manual labour Light manual labour Sedentary work Total
Number Operative result of patients Good Fair Poor
Mean score
16
1
7
8
84
6
1
3
2
97
5
2
1
2
11 4
3
1
2
97
30
5
14
93
11
pation, one had shifted to lighter work and 11 were disabled owing to back pain and one for another reason. Operative Results and Factors Influencing Them Assessed by the criteria adopted (Table 3), the result was good in five cases ( 17%), fair in 11 (37 %) and poor in 14 ( 47%). Among the patients under 40 years old, the result was at least fair in 75 % and poor in 25 % The corresponding figures for those over 40 were 29 % and 71 % (Table 4) The patients over 40 had statistically significantly (P< 0 025) poorer results than the patients under 40. A poor result was noticed in half the labourers in heavy work and in about one-third of those performing moderately heavy or light manual work preoperatively (Table 5) The difference was not statistically significant because of the deviation of the scores. The duration of symptoms was over 10 years in 14 cases The result was good in two of these, fair in four, and poor in eight The corresponding figures for 16 patients with back pain for less than 10 years were three, seven, and six The difference in operative results between these groups was not statistically significant.
T Tunturi et al : Lumbosacral Fusion by Standing H-Graft Technique Table 6 Operative results compared with preoperative duration of disability in 30 patients with radiologically successful fusions Preoperative disability dions (months)
Number Operative result of patients Good Fair Good Fair Poor
Mean score
1-6 6-12 > 12
6 13 11
2 3
4 5 2
5 9
15 2 10 6 45
Total
30
5
11
14
93
Table 7 Operative results compared with the number of previous back operations in 30 patients with radiologically successful fusions Number of Number Operative result previous back of patients Good Fair operations Good Fair Poor operations
Mean score
0 _1
14 16
2 3
8 3
4 10
11 4 74
Total
30
5
11
14
93
Table 8 Operative results compared with diagnoses in 30 patients with radiologically successful fusions Operative diagnosis Spondylolysis and spondylolisthesis Disc degeneration Sequel of previous disc excision Lumbarization-
Number Operative result of patients Good Fair Poor
Mean score
Of those 16 patients who had undergone previous back operations, nearly two-thirds had a poor result (Table 7) In this group the operative results were statistically significantly (P< 0 05) poorer than in the patients with no previous back surgery. The proportion of poor results was lower in the group of patients operated on for spondylolysis and spondylolisthesis than in the other groups (Table 8). In the group of one-segment fusions (20 patients) a good result was noted in four patients, a fair result in six, and a poor result in ten Of those ten patients who had undergone multiple level fusion, one had a good result, five a fair result, and four a poor result. Radiologically Unsuccessful Fusions Of those seven patients in whom fusion was radiologically unsuccessful, one showed non-fusion in the segment L 5-sacrum, to which the fusion was extended, although the segment under treatment was L 4-L 5 This patient had a good result of operation In the remaining six patients the result was poor, assessed on the principles adopted (Table 3) Two of these patients reported that their back had become somewhat better, while the symptoms were unchanged in three and back pain was aggravated in one One patient had shifted to lighter work and five were disabled owing to back pain The mean score for the patients showing non-fusion was 3.8 This was statistically significantly (P< 0 0025) lower than the score for the group of patients with successful fusions.
5
1
3
1
12 8
RadiologicalResults
10 13
1 3
5 2
4 8
10 6 73
30
5
11
14
93
Out of 40 patients attending the follow-up examination, 37 had been operated on by the standing H-graft technique Fusion was radiologically successful in 30 of these patients, while non-fusion was observed in seven. One-segment fusion was successful in 20 ( 95 %) out of 21 patients (in 94 % of patients immobilized for 1-3 weeks, in 100% of patients immobilized for six weeks). Multiple level fusion was successful in 10 (63 %) out of 16 patients One-segment fusions were statistically significantly (P< 0 025) more often successful than multiple level fusions.
sacralization
Unsuccessful previous fusion
Total
107
The operative results are compared with preoperative duration of disability in Table 6 The differences in results between all three groups were statistically significant (between first and second P < 0 005, second and third P< 0 0025, first and third P
