 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 1997 The Journal of Bone and Joint Surgery, Inc. Interlocking Intramedullary Nailing with and without Reaming for the Treatment of Closed Fractures of the Tibial Shaft. A Prospective, Randomized Study* , VANCOUVER, BRITISH COLUMBIA, CANADAInvestigation performed at Vancouver Hospital and Health Sciences Centre, Vancouver
One hundred and fifty-two patients who had 154 closed fractures of the shaft of the tibia were prospectively randomized to management with interlocking intramedullary nailing either with or without reaming. Thirteen patients who had been randomized to treatment without reaming were switched to the group that had reaming because of technical reasons; these patients were excluded from the analysis of the results. An additional five patients were lost to follow-up. Thus, seventy-two patients (seventy-three fractures) who had been managed with nailing with reaming and sixty-three patients (sixty-three fractures) who had been managed with nailing without reaming were available for follow-up at an average of twelve months (range, three to thirty-three months) postoperatively. The two groups were similar with regard to demographics and the configurations of the fractures. The average total duration of the procedures performed without reaming was eleven minutes shorter than that of the procedures done with reaming (p = 0.0013). The duration of fluoroscopy was not significantly different between the two groups (p = 0.35, Mann-Whitney test). The average estimated blood loss was identical for the two groups. Seventy fractures (96 per cent) that were treated with nailing with reaming and fifty-six (89 per cent) that were treated with nailing without reaming united without the need for an additional operation (p = 0.19). Because of the small sample size, the study has insufficient power (34.7 per cent) to detect this difference if it is real. There was only one deep infection, which developed after nailing without reaming. The nail fractured after one procedure with reaming. A screw fractured after two procedures with reaming and after ten without reaming (p = 0.012); multiple screws fractured after three procedures in the latter group. Malunion occurred after three nailing procedures with reaming and after two without reaming. Four malunions were of very proximal fractures and one was of a very distal fracture. Seventeen screws and twenty-four nails were removed after nailing with reaming, and twenty screws and nineteen nails were removed after nailing without reaming; neither of these prevalences was significantly different between the two groups (p = 0.27 and 0.89; chi-square test). We concluded that there are no major advantages to nailing without reaming as compared with nailing with reaming for the treatment of closed fractures of the shaft of the tibia. There was a higher prevalence of delayed union and breakage of screws after nailing without reaming.
Although many closed fractures of the tibial shaft can be treated non-operatively12,14,18, there are numerous indications for operative treatment. The preferred method of operative treatment remains controversial. Intramedullary fixation with a locking nail has become a popular technique because of reported excellent results in clinical comparative series2,4,6,7,10,11,21. The development of locking nails designed to be inserted without reaming has been stimulated by a number of factors. First, several studies have suggested that there is better preservation and more rapid recovery of the intraosseous blood supply after insertion of a small-diameter nail without reaming8,9,15,19,20. It has been inferred that fracture-healing should thus be improved compared with that after nailing with reaming and that infection should be less likely. Second, without the need for reaming, nailing should be simpler and quicker. Finally, given recent concerns with regard to fat embolism caused by intramedullary reaming of the femur and, to a lesser extent, the tibia, it has been suggested that nailing without reaming will lessen this complication13. Smaller-diameter nails may have some disadvantages. Failure of either the nail or the screws may be more common3. The smaller nails also may afford less stability to the bone and may thereby retard healing compared with that achieved with larger nails inserted after reaming. The purpose of this prospective, randomized study was to assess the clinical results of insertion of locking intramedullary nails with or without reaming for the treatment of closed fractures of the tibial shaft.
Beginning in October 1989, all patients who were seen at the Vancouver Hospital and Health Sciences Centre, a tertiary-care, university teaching hospital, for operative treatment of a closed fracture of the tibial shaft were considered for inclusion in the study. To be included, the patient had to be medically fit for an operation, could not have a concurrent infection, could not have had a previous injury with residual deformity or a previous local infection, and needed to provide informed consent for participation. The major component of the fracture had to be more than four centimeters distal to the tibial tuberosity and more than four centimeters proximal to the ankle. Thirteen fractures with an undisplaced intra-articular extension were included, with percutaneous fixation of the extension and nailing of the major diaphyseal component. After the patients had been entered into the study, they were randomized, with use of a closed-envelope technique, to treatment with nailing either with or without reaming. Patients who had bilateral fracture had each fracture randomized individually. The envelope was selected by the operating-room booking clerk. No patient refused to participate in the study. The study was approved by the Hospital Ethics Committee of our institution. Between October 1989 and April 1994, 152 patients (154 fractures) were entered into the study. Seventy-four patients who had a unilateral fracture were randomized to management with nailing with reaming, and seventy-six, to management with nailing without reaming. Of the two patients who had bilateral fracture, one had bilateral nailing with reaming and the other had nailing with reaming on one side and nailing without reaming on the other. Therefore, seventy-seven fractures were randomly selected for treatment with reaming, and seventy-seven were selected for treatment without reaming. Thirteen patients with unilateral fracture who had been randomized to management without reaming were actually managed with reaming for technical reasons; these patients were therefore excluded from the study, leaving sixty-four fractures in the group that had nailing without reaming. The procedure was switched to nailing with reaming because the canal was too narrow to accommodate the smallest (eight-millimeter) nail without reaming in six patients and because of a lack of a long enough nail designed to be inserted without reaming in four patients. In two patients who had a proximal fracture, the smaller nail designed for insertion without reaming did not afford good fixation of the proximal fragment; placement of a larger nail with reaming improved the stability. In one patient, insertion of a nail without reaming caused displacement of a butterfly fragment with tenting of the skin; after reaming, this no longer occurred. All of these fractures went on to heal without complications.
Operative Technique An attempt was made to perform the nailing within twenty-four hours after the injury; the limb was splinted in the interim. All nailing procedures were performed with use of a thigh bolster on a fracture table, calcaneal traction, and a Kirschner wire. The setup was identical for the two groups. Fluoroscopy was used throughout all of the nailing procedures. The reduction was achieved with use of the fracture table. After appropriate preparation of skin and draping, a midline incision was made over the patellar ligament. The approach (through the patellar ligament or medial to it) was selected according to the preference of the attending physician. The proximal part of the tibia was broached, and a guide-wire was passed. When the nailing was to be performed with reaming, the guide-wire was passed down the medullary canal, with an attempt made to center it in both the proximal and the distal fragment. When the nailing was to be done without reaming, the guide-wire was passed centrally only into the proximal fragment. Reaming was then carried out throughout the entire extent of the medullary canal in the former group and only in the proximal six to eight centimeters of the metaphysis, to accommodate the proximal flare of the nail, in the latter group. In the former group, the reaming was carried out to provide good fill of the medullary canal. The canal was overreamed to one millimeter more than the diameter of the nail. The appropriate length of the nails that were to be inserted after reaming was determined with use of the guide-rod subtraction method. The appropriate diameter of the nails that were to be inserted without reaming was determined by passing sounds of increasing diameter into the medullary canal; the diameter of the first sound that did not pass easily was selected as the diameter of the nail. The appropriate length of these nails was determined with use of a notched ruler under fluoroscopic inspection. The nails were inserted over a guide-rod in the group that had reaming and without a guide-rod in the group that did not have reaming. Insertion of the nail was controlled under image intensification. The locking configuration (static or dynamic) and the number of locking screws were chosen by the attending surgeon. Proximal locking was done with the aid of nail-mounted targeting devices, whereas distal locking was done with a freehand technique with use of a Pennig device (Howmedica, Rutherford, New Jersey). Postoperative immobilization was used only if it was needed for associated injuries or if the fixation was considered to be unstable. In the absence of injuries of the contralateral lower extremity, patients who were able to walk with a walker or crutches did so, without weight-bearing. Early motion of the knee and the ankle was encouraged. Weight-bearing was allowed at six weeks, depending on the progression of healing and the associated injuries. If consolidation of the fracture appeared to be delayed on serial radiographs, removal of the locking screws from the end of the nail most distant from the site of the fracture was considered. Also, if there was pain related to the prominence of the locking screws and it was too early for complete removal of the hardware, consideration was given to removal of the offending screws. Generally, a nail was removed only if it had been in place for at least a year and the patient had pain. Attempts were made to evaluate the patients at six weeks and then monthly until fracture-healing. Because of the large geographic referral base, some patients who lived far from our institution were managed by their family physician or a local orthopaedic surgeon. Demographic, injury-related, operative, and postoperative data were collected prospectively at the time of admission, the time of the operation, and the time of discharge from the hospital. Follow-up data were accumulated from office and clinic records for local patients and through the family physician or the local orthopaedic surgeon for patients from other communities. All radiographs were assessed with regard to the pattern, location, comminution, displacement, and healing of the fracture. The patients were followed at least until the fracture had united or until a non-union had been treated definitively. The duration of follow-up averaged twelve months (range, three to thirty-three months). The fracture was considered to have united if the patient was able to bear full weight without pain at the fracture site and if radiographs showed callus-bridging at the fracture site or obliteration of the fracture line in patients who did not have major formation of callus. A non-union was diagnosed if serial radiographs, made over a period of at least three months, showed no progression toward healing. Malunion was defined as angulation of more than 5 degrees as seen on anteroposterior or lateral radiographs, more than one centimeter of shortening or lengthening, or more than 15 degrees of rotational malalignment. Statistical analysis was carried out to compare the two groups (nailing with reaming and nailing without reaming). A Student t test was used for normally distributed, ordinal-scale variables. A Mann-Whitney two-sample t test was used for non-normally distributed, ordinal-scale variables. Either a Fisher 2 x 2 exact test or a chi-square analysis was used for data on a nominal scale. The difference between the two groups was considered to be significant when p < 0.05.
Data on the Patients With the numbers available, no significant differences were detected between the two groups with regard to the level or the pattern of the fracture, the grade of comminution according to the criteria of Winquist and Hansen, or the degree of displacement (Table I).
Eight fractures that were treated with nailing with reaming and five that were treated without reaming were associated with a fracture of the ipsilateral ankle or tibial plateau. These were undisplaced fractures that were treated with percutaneous screws. The average interval between the injury and the operation was seventeen hours (range, three to forty-eight hours) for the group that had nailing with reaming and twenty-two hours (range, five to 112 hours) for the group that had nailing without reaming. With the numbers available, this difference was not significant (Mann-Whitney test, p = 0.13). The operative times for insertion of the nail, proximal locking, and distal locking, as well as the total time from the initial incision to the closure of the wound, were assessed. With the numbers available, significant differences were found only for insertion of the nail (p = 0.0001) and total time (p = 0.0013) (Table II). The nailing procedures that were performed with reaming required, on the average, eleven more minutes of operative time than those performed without reaming.
The duration of fluoroscopy, although not a very reliable item of data because of use of various machines with different modes of exposure, was not significantly different, with the numbers available, between the two groups (average, one minute in both groups; p = 0.35, Mann-Whitney test). The nails that were inserted after reaming had an average diameter of eleven millimeters (range, nine to thirteen millimeters), compared with an average of nine millimeters (range, eight to eleven millimeters) for those inserted without reaming. The length of the nails was similar in both groups (average, 336 millimeters [range, 270 to 405 millimeters] for the nails inserted after reaming, compared with 338 millimeters [range, 285 to 360 millimeters] for those inserted without reaming). A Grosse-Kempf nail (Howmedica) was used in all procedures that were performed with reaming and in two that were performed without reaming. A Delta nail (Smith and Nephew Richards, Memphis, Tennessee) was used in the remaining procedures that were done without reaming. Seventy-three (95 per cent) of the nails inserted after reaming and sixty-two (97 per cent) of those inserted without reaming were locked both proximally and distally (the static configuration). In the group that had nailing with reaming, one proximal locking screw was used for sixty-six fractures; two, for seven; and none, for four. One distal locking screw was used for forty-four fractures and two, for thirty-three. In the group that had nailing without reaming, one proximal locking screw was used for fifty-seven fractures; two, for six; and none, for one. One distal locking screw was used for forty fractures; two, for twenty-three; and none, for one. The average estimated blood loss was 160 milliliters in both groups. In both groups, the patients who had an isolated injury stayed in the hospital for an average of five days (median, four days) and the patients who had multiple injuries stayed for an average of nine days. Follow-up data were not available for four fractures that were treated with nailing with reaming and for one that was treated without reaming; thus, seventy-three fractures (95 per cent) that were treated with reaming and sixty-three (98 per cent) that were treated without reaming were available for follow-up through the time of fracture-healing.
Perioperative Complications Other perioperative complications after nailing with reaming included deep venous thrombosis, arrhythmia, fat embolism, and pulmonary embolism (one patient each). The patient who had a fat embolism had multiple injuries, including fracture of the ipsilateral femoral shaft. In the group that had nailing without reaming, two patients had pneumonia and two had deep venous thrombosis. All of these complications resolved after medical treatment.
Union Seven patients had a non-union after nailing without reaming. Six of them had revision internal fixation (repeat nailing in five and application of a plate in one) at an average of nine months (range, seven to fourteen months) after the injury. Four of these six patients had not had union after the original nailing despite conversion of the fracture to a dynamic status, but all four had union after the revision. The seventh patient still had a non-union at the most recent follow-up evaluation, at nine months, and additional assessment and treatment were planned. The difference in the rates of non-union, with the number of patients available for study, suggested a trend but was not significant (p = 0.19, Fisher exact test).
Infection
Failure of the Implant Two nailing procedures that were performed with reaming and ten that were performed without reaming were complicated by fracture of at least one screw (Table II). Three procedures in the latter group were followed by fractures of multiple screws. This difference was significant (p = 0.012, Fisher exact test). A screw backed out from four nails that had been inserted with reaming and two that had been inserted without reaming. With the numbers available, this difference between the two groups was not significant (p = 0.69, Fisher exact test) (Table II).
Malunion
Subsequent Procedures The nail was removed from the sites of twenty-four (34 per cent) of seventy fractures that united after nailing with reaming and from the sites of nineteen (34 per cent) of fifty-six that united after nailing without reaming. With the numbers available, this difference was not significant (p = 0.89, chi-square test). Seven patients who had had nailing with reaming had other, subsequent operative treatment. Two had revision nailing for a non-union; two, excision of painful ununited fragments (one had bone-grafting at the same time); and one, an early derotation (revision distal locking) for a rotational malalignment noted in the early postoperative period. Another patient was seen two weeks after the injury, after having had an uneventful postoperative course, with an impending compartment syndrome. At the time of the operation, a false aneurysm of the anterior tibial artery was noted. This lesion was subsequently repaired, and fasciotomies were performed. The seventh patient had a shortening osteotomy because of a severe open fracture of the contralateral tibia with bone loss. Eight patients who had had nailing without reaming had other, subsequent operative treatment. Four of these patients had revision nailing for a non-union; one, an osteotomy and application of a plate for a fracture that appeared healed radiographically with malunion but that was found to be ununited at the time of the osteotomy; one, operative treatment for an infection; one, a heel-cord lengthening; and one, an exploration of the extensor hallucis tendon for tendinitis that was possibly related to exuberant callus.
The results of interlocking nailing with reaming in our series are in keeping with those reported by others2,4,7. Data on the outcome of interlocking nailing without reaming for closed fractures are limited, but reports of its use for open fractures have revealed high rates of failure of screws and non-union16,17,21. Riemer et al. reported the results of interlocking nailing without reaming in a series of sixty-seven patients, including twenty-six who had a closed fracture. Similar to our experience, they found no malunions associated with more than one centimeter of length discrepancy or more than 10 per cent angular or rotational malalignment. There were no infections, but seven (27 per cent) of the twenty-six patients needed exchange of the nail or application of a plate with bone-grafting for delayed union or non-union. One nail and four screws broke in the sixty-seven patients. The nails that were inserted without reaming were different from those in our study, perhaps explaining the difference in the rates of failure of the implant. Our results do not appear to justify use of a nail without reaming in the treatment of a closed fracture of the tibial shaft that necessitates operative fixation. The only advantage is that the operative time is slightly (average, eleven minutes) shorter than that for nailing with reaming. Nailing without reaming had technical limitations, which were partially related to design and inventory factors. Despite the theoretical advantage of better preservation of the blood supply8,9,15,19,20, nailing without reaming does not seem to be associated with a better rate of union. In fact, the results suggest a trend toward a higher rate of non-union. It may be hypothesized that this is related to the less stable fixation that is provided by the smaller nail and perhaps to the beneficial effect of the material produced by the reaming acting as a bone graft. It is interesting that the fate of the material produced by the reaming was unpredictable. Although this finding could not be quantified, the serial radiographs of some patients showed that material produced by the reaming had disappeared, whereas those of others showed that the material had been incorporated into the callus. The subjective impression was that this depended partially on the rigidity of the fixation. As might have been anticipated, infection did not prove to be a problem in either group of closed fractures; therefore, any considerations regarding preservation of the blood supply in the prevention of infection appear to be theoretical only. Failure of the smaller nails that are inserted without reaming is a concern. The weak link is evidently the screws, as the nails showed no tendency to fail. The screws, probably because of their smaller size and their different design, were significantly more likely to fracture after nailing without reaming than after nailing with reaming (p = 0.012). It is conceivable that the locking screws used in nailing without reaming could be modified to reduce the prevalence of failure. Fracture of screws had no detrimental effect in this study, although anecdotally we have encountered difficulty in removing intramedullary nails when screws are broken. This is primarily a problem with fully threaded screws, which may fail at the nail-screw junction closest to the screw head, thus complicating removal of the nail. There were numerous examples, again not quantifiable, in which failure of screws appeared to aid in fracture-healing by shortening the fracture gap and loading the fracture site. Perhaps earlier and more frequent conversion of the fracture to a dynamic status is indicated both to encourage healing and to prevent the failure of screws. There were no apparent complications related to the reaming. In fact, with one exception (a patient who had a delayed compartment syndrome in association with a traumatic aneurysm), compartment syndromes occurred only in the group that had nailing without reaming. This clinical experience is in keeping with unpublished data from our center that showed no substantial sustained increase in compartment pressures in association with reaming. The only fat embolism in the current series was in a patient who had multiple injuries, including multiple fractures of long bones. The relationship between reaming and fat embolism remains to be defined. The data in this study do not support use of a nail without reaming to avoid fat embolism. We do not share the enthusiasm for nailing without reaming in the treatment of closed fractures of the tibial shaft. Although insertion without reaming shortens the duration of a nailing procedure, the operation has technical limitations. There is a tendency toward an increased rate of non-union and a higher rate of failure of the screws.
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
  CiteULike  Connotea  Del.icio.us  Facebook  Technorati  Twitter What's this?
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
| home | CME | help | feedback | subscriptions | archive | site index |
|
Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online. |