U.S. Army Medical Department, Office of Medical History
Skip Navigation, go to content







AMEDD MEDAL OF HONOR RECIPIENTS External Link, Opens in New Window






Chapter V







During the four-year period ending December, 1923, 129 bone-graft operations were performed at Walter Reed General Hospital for conditions resulting from World War wounds followed by severe infection of bone and soft parts, for nonunion, and loss of substance in bones of the extremities. Fifty-two of these were unsuccessful. The majority of these cases were the result of war wounds in which, at the time of injury, severe damage was sustained by the bone as well as its surrounding soft parts. The bone showed marked atrophy and osteoporosis; its osteogenetic power was at a minimum, latent infection was present in the bone and the surrounding sear tissue, and the circulation in both was markedly impaired. In practically till of these cases the fractures were received during the summer of 1918: some even earlier. The majority occurred in France and had active infection in the bone for from four months to two or more years, causing a destructive osteitis, the usual infecting organism being the hemolytic streptococcus. Ordinarily this condition of chronic bone infection has been referred to as osteomyelitis, from which it differs in many respects, but by usage has come to mean the same. Repeated operations for the removal of sequestrum, establishment of drainage, and preparation of the wound for dakinization, further limited the blood supply, increasing scar tissue formation and bone atrophy. Added to this was the atrophyof disuse which further impaired its reparative osteogenetic properties.

As other Army general hospitals closed, many of the failures there, as well as the nonunion cases that were still septic, were gradually transferred to Walter Reed General Hospital, the easier and more successful cases having been cured and discharged. Others were sent in for treatment by the Veterans' Bureau.

In this group, 26 had been grafted elsewhere unsuccessfully; 15 had been plated or wired, and 6 had had some type of "stepping" operation. For a 2-inch loss of substance in a radius, one surgeon had used a toothbrush handle unsuccessfully. Atrophy was probably most marked in the humerus and next in the tibia, and usually where there was loss of bone substance the bone ends were atrop1hic, rounded off, or pointed. Eburnation in bone ends with pseudo-arthrosis occurred in the tibia and bones of the forearm, without loss of substance, and occasionally in the femur and humerus with loss of substance, since there was no second splinting bone to hold the ends apart.  This was

a The data in this chapter are based on "End Results of One Hundred and Fifty-eight Consecutive Autogenous Bone Grafts for Nonunion In Long Bones (a) in Simple Fractures; (b) In Atrophic Bone Following War Wounds and Chronic Suppurative Osteitis (Osteomyelitis)," by Maj. N. T. Kirk, Medical Corps, U. S. Army. The Journal of Bone and Joint Surgery, Boston, 1924, vi, No. 4, 760-799


more often the case in the femur than in the humerus, as bone-end apposition was more frequent.

Loss of substance was the rule in this group, the amount varying from a fraction of an inch to 5 inches. The whole shaft of the humerus except 2 inches at each end was destroyed in eight cases. Several cases with 3 inches loss of substance in the tibia were grafted successfully. In a case having a 5-inch loss of substance, the head of the tibia formed the proximal fragment. This case was grafted once elsewhere and twice at Walter Reed General hospital, all three operations being unsuccessful: the first due to infection and in the latter two operations atrophy and fracture occurred. Eventually the leg was amputated.

No attempt was made to graft in any case until it had been healed for at least six months without signs of infection, unless the roentgenogram was negative for sequestrum or evidence of infection in bone, and only after vigorous repeated massage in the physiotherapy department. If scar tissue was present in skin, it was removed and a skin closure done, at which time any sear tissue in soft parts or about the bone ends was removed. It was found that scar tissue in skin in these cases would invariably break down if at all in proximity to the operative field, and often cause exposure of bone and disaster. If infection followed the sear excision, the graft was not attempted for another six months. If healing was by primary intention, the graft was done four to six weeks later. Even with this procedure, infection caused 22 failures, and 9other cases were severely infected, but union occurred, although the graft was later removed in 3. Very severe infection causing failure has been encountered after the original wound had been healed one year and no reaction occurred after a preliminary scar excision. There appears to be no assurance as to when all danger from latent infection is past. In the leg it was sometimes necessary to do two scar excisions and skin plastic operations before all scar tissue was removed and there was sufficient healthy skin to cover the bone.


The types of grafts included inlay, intermedullary, osteoperiosteal peg (not intermedullary). Grafts of the inlay type included true inlays, outlays. fish-tail'' type, and “massive" grafts. Most of the grafts were cut with the single Albee saw from the healthy tibia. In 1920 and 1921 the crest of the tibia was used in the humerus, radius, and ulna, hut this was later abandoned and the graft was taken from the inner surface. The inlay type of graft was always used in the tibia, usually in the forearm, occasionally in the humerus and femur.

The intermedullary type was found best suited in the atrophic humerus. The cortex was so atrophic that there was practically nothing left but the medulla, in which there was an increase in fat. The ends of the bone fragments were cut off and the graft introduced, causing a minimum of interference to the blood supply of the bone by way of the periosteum attached to soft, parts. Again, the graft was driven into the medulla of the upper fragment and inlaid into the bone of the condyles and supracondylar region. These


grafts were reinforced by osteoperiosteal grafts wrapped about them and sutured to the periosteum of the upper and lower fragments.

The osteoperiosteal type was used successfully on a fractured patella with nonunion, of four years' duration the result of a gunshot wound. The first attempt in this case was unsuccessful, due to the failure of the absorbable suture to maintain fixation of fragments sufficiently long. A second failure occurred in the use of this type of graft on a fractured graft. The size of the graft and its contact with healthy bone is essential to its circulation, life, and proliferation. The general rule followed is to cut a graft at least three times the length of the loss of substance to be bridged or of poor bone in which it will be in contact, though it is not always possible to get a graft long enough from the healthy tibia to meet these requirements.b At first these grafts were held in place with kangaroo tendon and chronic catgut through drill holes in the side of the trough. It was found, however, that the circulation was so poor that this absorbable material did not absorb but acted as an irritant, causing sinus formation, and was more than once the cause of infection and loss of the graft. The writer has removed it as long as one year after it was placed in the bone. Its use was discontinued and the grafts made self-retaining without the use of ligatures. This was accomplished by cutting the graft to fit snugly, and a half inch longer than the trough in the inlay type, undercutting the ends of the cortex on both fragments and sloping the graft ends from above downward so that they could he wedged under the cortex at both ends. The graft is then fitted in position with one end under the cortex, forced down to position at the other, and then, using a mallet and an instrument with a sharp point and a shoulder, slid down until the lower end becomes fixed under the cortex. This method is now being used in all grafts.

There were eight cases in this series of bone grafts of the tibia operated upon by an associate of the writer in which an entirely different procedure was used; all were successful. He used a small graft cut from the inner surface of the tibia and, after removing all of the endosteum, secured the graft into the fragments without opening the medulla. The graft consisted of periosteum and osteum only and was placed in contact and made self-retaining in the cortex by cutting a wedge, sloping the graft ends from above downward. and sliding the graft in from the side, the end becoming engaged under the notch that had been cut in the cortical bone. Chances of infection were lessened because there was less bleeding, the medulla not being opened; bone growth was slower than in the inlay type, but occurred. He applied his plaster cast before operation and operated through a window in the plaster.

The tourniquet was not used except in operating upon both bones of the forearm when the bloodless method was employed, but it was removed and all bleeding controlled before the graft was put in position. Hemostasis was difficult in these chronically infected cases, due to oozing from scar tissue. but was as complete as possible before closure.

The skin, except when two incisions were necessary in the forearm, was sutured with silkworm gut, a window was cut in the plaster cast two days
b The average size of inlay graft used in the tibia for nonunion was 6 by one-half inch; the largest measured 10¾ by three-fourths inch, in a tibia with 3-inch lost substance.


after operation, and the wound dressed daily until all stitches were removed when the window was filled in with plaster. This was necessary because of possibilities of infection. When infection occurred, stitches were removed and the wound at once dakinized. Nine cases in which infection was severe were saved by this method and many small local skin infections Were controlled early and severe infection avoided. Plaster was removed at the end of three months in the tibia and forearm, a roentgenogram taken, and plaster reapplied for another three months in the chronic healed osteomyelitis group. Some required immobilization even longer than this. A well-padded body cast was applied at least two days before a humerus was grafted to insure its proper fitting and setting. The arm piece was put on after the graft was completed. This afforded more comfort to the patient, gave better fixation, and lessened the time on the operating table. In a few cases the arm piece as well as the body cast was applied before the operation and the operation done through a window. This made the procedure much more difficult and was finally abandoned.

The arm was put up in abduction unless the nonunion was between the pectoral and deltoid insertions, when it was adducted and brought across the chest for fear of fracture of the graft from muscle pull.

These casts were not removed until the end of six months.

The following tal)le shows the bone grafted and the results:


From this table it will he seen that there were six patients in whom two bones were involved; five involved the forearm, the sixth was a patient having an old gunshot wound with compound comminuted fracture and incomplete union in both tibiae, lower third. One tibia had previously been grafted elsewhere and had fractured. Both legs presented adherent scars. The patient begged for a double amputation. This was refused. Four operations were done, one at a time, two sear excisions and two inlay grafts taken from the upper third of the bone being grafted. Both were successful. In addition, the patient had bilateral drop-foot, due to loss of muscle and tendon. One side was corrected by tendon suture at the time of the bone graft. In the other the tendons were hopelessly destroyed. The patient is now walking without laces, except a light one to correct the foot-drop in the left leg.

There were two cases in which the shoulder joint, head of the humerus, and from 2 to 3 inches of the shaft had been shot away, along with the deltoid


muscle. In one case the acromion process, all outer end of the clavicle were missing. The nerve ind blood supply to the arm had not been disturbed. Along adherent sear replaced the deltoid muscle and the long head of the biceps. The arms were useless . After anchoring the remaining shaft of the humerus to the glenoid cavity, at an angle of 90 degrees abduction and in a neutral position as regards flexion anil extension, an excellent functioning result was obtained by use of scapular motion in both cases. In the first case, after cleaning out the glenoid cavity, cutting, off the end of the humerus, the acromion process was incompletely fractured (green-stick), brought down, and a peg from the crest, of the tibia was driven through a hole bored in the acromion, the humeral shaft, and into the glenoid. In the second the glenoid was cleaned out and the humerus cut off and fitted to it. A hole was made 1 inch deep and three-eighths inch square in the glenoid. A 4-inch graft was cut from the tibia, driven into what remained of the medulla of the humerus for 3 inches. The humerus with the protruding 1 inch of graft was then fitted to the glenoid with the hole that had been prepared to receive the graft. The humerus was then anchored to the scapula with a piece of silver wire.

One forearm had nonunion of both bones with marked deformity, pseudo-arthrosis, and severe eburnation of the bone ends, with two large adherent scars.The scars were excised, both ends of the bone were resected, and later, using an osteoperiosteal graft on the radius and an interimedullary peg in the ulna.union occurred. In another case there was malunion in the ulna with bad deformity in two directions, and nonunion in the radius, which contained piece of broken silver wire. The ulnar deformity was corrected by osteotomy and an intermedullary peg, the wire was removed, and a 5-inch inlay self-retailing graft placed in the radius. Excellent bony union in both, correcting the deformity.

Six times the nonunion in the raidius was in the lower end, presenting the characteristic radial deviation deformity. This required an osteotony and shortening of the ulna from an inch to an inch and a half to correct the deformity.

The smallest graft in the series was 114 inches long and three-sixteenths of an inch square, and was used to replace the shaft of the second metacarpal bone. The result was excellent. The tendons had not been destroyed and after a capsulotomy posteriorly of the metacarpophalangeal joint, the soldier was returned to duty with normal function.

One tibia had grafted with a heterogenous graft, taken from the tibia of another patient of the same blood group, after amputation through the middle third of the leg. This was used because of an old healed chronic osteomyelitis in the opposite tibia, and a sliding graft was impossible on the one having the nonunion, there not being sufficient bone. Absorption and infection followed the osteoclasts appeared to be very active.

One tibia had malunion with a large unhealed scar 3 inches in diameter. The lower fragment was displaced outwardly 20 degrees from the long axis of the upper fragment in the position of weakness for weight bearing. The scar was cauterized with the actual cautery, excised, a correcting osteotomy was performed on the tibia, but there was no attempt at union: some of the bone sequestered After another sear excision, tile tibia was successfully grafted


the deformity being corrected. Fifteen cases were grafted a second time. Seven were successful; three of these had been unsuccessfully grafted elsewhere, making their third graft. Four were tibias, one was an ulna one a patella, and one a humerus.

Of the eight failures, five had been unsuccessfully grafted elsewhere before, this making the third graft.

The tabulated results of the total series are:

    Percent.............................................. 59
Unknown..............................................   1
Failures................................................. 52

Only cases which had definite bone union and had a functional result on discharge were classified as successful. Patients were not discharged until after there was sufficient bony union in the lower extremity to allow them to walk with the use of a brace only to guard agaist undue stress, and those of the upper extremity were held for a like period. This period was anywhere from 6 to 20 months after the graft.

Union was never sufficient to allow weight bearing before the sixth month and in some cases not until the ninth and then only with properly fitting braces which were made in our own orthopedic shop. The war demonstrated that gunshot fractures which united without graft required twice the time to form solid bony union as was the case in peacetime fractures, and required splinting bracing for a like period to prevent refracture or deformity.

The same observation was made in this series of grafts: the period of time required for complete union to occur was from two to three times as long in the chronic osteomielytis group as in a noninfected group.

Two cases that were originally classified as failures due to infection subsequently returned for reexamination, when solid bony union was found. It is not impossible that a certain number of cases reported it here as failures have united since discharge.

The one case carried as unknown left the hospital in plaster and has not been heard of since.

The cause of failure was:
Fracture in plaster before seventh month........... 7
Fracture after seenth month.............................. 6
Faulty fixation of patella....................................1
Death from shock.............................................1

Infection was the cause of the greatest number of failures due, no doubt, to latent infection in the bone and soft parts, rather than to accidental infection at the time of operation. The same technique was employed and the same operators operated upon these cases, as in like operations in simple fractures. yet there were no grafts lost from infection in the simple fractures.


The cause, then, is considered to be latent infection, diminished blood supply in the bone and soft tissues, and the presence of subcutaneous scar tissue and bone atrophy, thereby lowering the vitality of the part. Sequestra not infrequently formed at the end of the fragments along the edge of the trough; apparently the interference with the circulation of the bone in cutting the trough caused its death.

Two sets of instruments were always used when a graft was done, one for the extremity being grafted, and the other for the healthy tibia. Infection was carried into the tibia of the healthy leg once by the use of the same twin saw. Its use was promptly discontinued.

Atrophy was apparently the cause of 15 failures. This may have been the result of using too small a graft with insufficient bone contact to nourish it, improper fixation, or insufficient blood supply and ostpogenetic power in the bone which was grafted to keep the graft viable. The failures actually occurred in those cases showing most marked bone atrophy or loss of substance and in cases with some underlying constitutional disease such as tuberculosis and syphilis, or were in generally poor physical specimens.

Fracture of the graft before the seventh month caused failure in seven cases. Atrophy undoubtedly played a part in these. Other factors were muscle action, poor fixation due to atrophy of the soft parts and shrinkage of the extremity in its cast, and to too much activity, carelessness, and non-cooperation on the part of the patient.

There were nine fractures after the seventh month while wearing braces, some of these as late as the ninth month, all due to trauma. All originally had loss of substance of from 1 to 3 inches. All these might readily be classified as successful. One patient discarded his brace and apparently deliberately refractured his graft, as he did not desire his discharge. In three cases the bone reunited without operation, leaving six failures due to this cause.

Two patients were reoperated upon successfully with inlay grafts along the fractured graft; one unsuccessfully, using an osteoperiosteal graft; one continued to run a systolic blood pressure of from 80 to 100 and was refused further operation. Two left the hospital without further operation; one developed a sinus after three months, and the graft was lost through infection the eighth month, due, it is believed, to irritation from the kangaroo tendon ligature.

One failure in a grafted patella occurred, due to absorption of absorbable suture used to hold the fragments in position until consolidation occurred after an osteoperiosteal graft. This case was regrafted, the fragment being held with silver wile. Union occurred, and the knee flexes to 900 and the patient has sufficient power in the quadriceps to go up and down stairs. There was one death from surgical shock in a graft of the upper third of the femur.

A blood pressure reading was taken during all grafts every 10 to 20 minutes, and a careful check kept on the patient's condition. Shock was much more easily produced than is ordinarily the case, due to the long hospitalization, with absorption of toxins from chronic infection and repeated


operations. When it occurred it was treated with saline intravenously and blood transfusion. A word about the tibia from which the grafts were taken. As was stated, infection was carried into the healthy tibia once. When the graft is taken from the inner surface, the defect is soon filled in with new bone. The writer has removed a second graft, 8 by one-half inches, six months after the first; the cortical done did not appear as well calcified, but was thicker than normal. He has removed a third large graft from the same tibia and still found good hone.

When the crest is removed, however, it is not wholly replaced and the patient can easily feel the bone defect with his hand.

All patients, when they again became ambulatory about the third week after operation, were fitted with a tibial caliper which was worn for from 8 to 10 weeks, depending upon the amount of bone removed. Two fractures occurred in this series; one patient, not wearing his brace, got too close to a motor truck, and another fell down a flight of stairs while intoxicated. Both healed without deformity.


CASE 1.- A. H. Gunshot wound, left leg, causing compound comminuted fracture, and loss of .3 inches of substance in left tibia, followed by chronic osteomyclitis and marked bone atrophy. A large part of the musculature, as well as of the skin, was destroyed. Two scar excisions and plastic skin closures were necessary before a graft could be attempted. Operation, December 2, 1920, Walter Reed General Hospital. Bone graft 7 inches by one-half inch was taken from the right tibia and inlaid into both fragments and made self-retaining. The graft was covered on three sides at the site of nonunion with skin only, and a small sinus developed, which required operation in July, after which the wound healed. Patient was fitted with ischial caliper in December, 1921. He was discharged October ,1922, with good bony union.
CASE 2.- N. W., Pvt., Inf. Gunshot wound of right forearm, causing compound corn-minuted fracture, nonunion both radius and ulna at junction of upper and middle third, followed by chronic osteomyelitis. There were two large adherent scars, marked deformity, and eburnation of both ends of both bones. Operation, March 5, 1920, Walter Reed General Hospital. Scar tissue excised from skin; the eburnated ends of both bones excised and deformity corrected. A Lane plate placed on radius to keep position. This was removed August 4, 1920, when fibrous union was found to be present. September 23, 1920, operation Walter Reed General Hospital. Osteoperiosteal graft wrapped about the point of nonunion in radius and an intermedullary peg placed in the ulna, both taken from the left tibia. Excellent bony union occurred in the radius and fair union in the ulna. The patient was playing ball with this arm when discharged.

CASE 3.- C. B. H. Gunshot wound, right leg, causing compound comminuted fracture and loss of substance in tibia, followed by chronic osteomyelitis. Three-fourths of the breadth of the shaft in the middle third of the tibia was lost. Operation, May 4, 1921, Walter Reed General Hospital. Bone graft 9 inches by one-half inch was taken from the inner surface of the left tibia and inlaid into the normal bone of the upper and lower fragments and outlaid along the 6 inches of the remaining bone in the middle third. Healing occurred by first intention. Immobilization in plaster until February, 1922. Fluctuation was at this time found present and a bloody fluid was removed. The fluctuation recurred and the wound had to be incised and Dakinized. In July, 1922, a skin suture was done, and it healed by first intention. The bone was not infected. Patient discharged with good union, September 6, 1922.
CASE 4.- E. R. Gunshot wound, with compound comminuted fracture and loss of substance, both bones upper fourth of left leg, followed by chronic osteomyelitis. In addition, patient had paralysis of the external popliteal nerve. On admission the upper end of both


FIG. 85.- Case 1. Loss of bone substance and bone atrophy

FIG. 86.- Case 1. Roentgenogram 3½ months after graft


FIG 87.- Case 1. August 1, 1922. Roentgenogram shouting excellent bony union


FIG. 88.- Case 1. Roentgenogram, May, 1924, showing hypertrophy of graft in tibia

FIG. 89.- Case 2. Marked deformity and eburnation of bone ends where the pseudarthrosis had occurred


FIG. 90. - Case 2. After resection of bone ends and removal of plate. Deformity corrected

FIG. 91.- Case 2. March 23,1921. Excellent bony union and hypertrophy of radius


FIG. 92.- Case 3. Roentgenogram, December 10, 1921, showing bone being thrown across between graft and old eburnated bone

FIG. 93.- Case 3. Roentgenogram, July 28, 192, 14 months after operation. There is excellent bony union


lower fragments of the tibia and fibula were in apposition with the lower end of the head of the tibia. No attempt at union occurred. Operation, August 31, 1921, Walter Reed General Hospital. Bone graft taken from inner surface right tibia was driven into the remaining head of left tibia and inlaid into the upper end of the lower fragment of the tibia. A small slough occurred in the suture line, which healed. Plaster was removed the seventh month and an ischial caliper substituted. Patient discharge one year after graft with excellent bony union in the tibia, as well as union between the head of the tibia and the shaft of the fibula, which occurred only after the tibia was grafted.

FlG. 94.- Case 4. Condition before operation

CASE 5.- W. J. T. Gunshot wound, with compound comminuted fracture, lower third both tibia and fibulae, followed by chronic osteomyelitis and loss of muscle substance from extensor muscle group in both legs. Before admission patient had had a bone graft in right tibia, which had fractured. He requested that both legs be amputated. This was refused. A scar excision was done on each leg, at Walter Reed General Hospital December 15, 1921, a bone graft 5 inches long was taken from the upper third of the same tibia and inlaid in the lower third alongside old fractured graft, and made self-retaining. A tendon suture of the right extensor tendon group was done at the same time. February 6, 1922, the left tibia was


grafted, using a graft 4½ inches by one-half inch, taken from the upper third of the same tibia, and made self-retaining. Healing of skin was delayed in both cases but no infection occurred. Excellent bony union occurred after each graft.
CASE 6.- D. A. H. Gunshot wound, with compound comminuter fracture, and loss of head, neck, and upper 2 inches of humerus, as well as deltoid muscle, followed by chronic osteomyelitis. The blood and nerve supply to the rest of the arm was not disturbed. The arm hung uselessly by the side. Operation, October 13, 1921, Walter Reed General Hospital. Body cast was applied two days before operation. Arthrodesis left shoulder. Glenoid was cleansed out, the acromion process was incompletely fractured (green-stick) and brought down; the end of the humerus cut off and fitted to the glenoid in a position of 90 degrees abduction and neutral as to flexion and extension. A bone peg from the crest of the tibia was driven through a hole that had been drilled in the acromion process through the upper end of the shaft of the humerus and into the glenoid, fixing the humerus in position. The armpiece of the cast was applied. The cast was removed at the end of six months and good bony union was found. The arm was kept on an airplane splint for four months. Patient discharged with good scapular function.
CASE 7.- W. H. M. Gunshot wound, with loss of head and upper fourth of shaft of humerus, acromion process, and outer end of clavicle. This was followed by chronic osteomyclitis. The deltoid muscle and long head of the biceps was replaced by scar tissue; the circulation and musculature of the rest of the arm was normal. The arm hung uselessly by the side. Operation, November 9, 1922, Walter Reed General Hospital. Arthrodesis of left shoulder. The glenoid was cleaned out; the upper end of the humerus sawed off. A graft 4 by 3.8 inches was taken from the inner side of left tibia and fitted in a hole 1 inch deep made in the glenoid; removed and driven into the medulla of the humerus. The humerus, with the protruding 1 inch of the graft, was then fitted to the glenoid and its cavity. Through a drill hole in the shaft of the

FIG. 95.- Case 4. Solid bony union January 17, 1922, five months after graft

humerus and the coracoid process, a silver wire was passed, securing the humerus to the scapula. Immobilization in plaster with the humerus at 90 degrees abduction and in a neutral position as regards flexion and extension. Body cast was applied two days before operation and armpiece at time of opera-ion. Five months after arthrodesis, plaster was removed and excellent union was found to be present. Discharged the ninth month with excellent scapular motion.
CASE 8. L. E. Gunshot wound, sustained in action, causing compound comminuted fracture and loss of substance middle third right humerus, and paralysis of musculospiral nerve. There was also a chronic osteomyelitis. Before admission humerus was "stepped"


and a 4-inch bone peg introduced into medulla, followed by nonunion. Operation, April 20,1922, Walter Reed General Hospital. Intermedullary bone graft 6 by ½ inches taken from the right tibia, made self-retaining in medulla. Cortex of fragments had consistency of an eggshell. Cast removed at end of six months, wound having healed by first intention. Roentgenogram showed good bony union. Discharged July, 1923.

FIG. 96.- Case 5. Anteroposterior view of both tibiae before bone graft

CASE 9. O. P. Patient sustained, along with other injuries, a gunshot wound causing compound comminuted fracture of the right patella. On admission in August, 1922, there was union between the two upper fragments of the patella and separation of some 2 inches between the upper and lower fragments. Knee flexed to about 450; muscle power in quadriceps was poor. Operation, September 1, 1922, Walter Reed General Hospital. Scar tissue was removed from between fragments, which necessitated opening knee joint; patella


FIG. 97.-  Case 5. Right tibia four months

FIG. 98.- Case 5. Left tibia four months after graft


was held in position with chromic catgut, using a circular purse-string suture anti suturing the lateral capsule. An osteoperiosteal graft was placed over and in contact with the anterior bony surface of both fragments. The cooperation of this patient was poor. The absorbable ligatures did not hold and the fragments separated.
Reoperation, December 4, 1922, Walter Reed General Hospital. The same procedure was followed as in the first operation, except that the patella was drilled and two pieces of silver wire used for fixation through the drill holes. A new osteoperiosteal graft was

FIG. 99.- Case 6. Roentgenogram showing loss of bone substance before operation

placed in contact with the anterior surface of the patella and a third piece of silver wire run through the patella tendon and the insertion of the quadriceps tendon, reinforcing the fixation. Excellent bony union occurred, and when the patient returned in January, 1924, to have a piece of the silver wire, which was broken, removed, he had solid bony union, 90motion, and sufficient power in his quadriceps, to go up and down stairs.
CASE 10. A. B. L., private, Infantry. Gunshot wound of right hand causing compound fracture of second, third, and fourth metacarpal bones, followed by mild infection and loss of shaft of second metacarpal July, 1922. Operation, 'March 21, 1923, Walter Reed General


FIG. 100.- Case 6. Good bony union six months after arthrodesis

FIG. 101.- Case 6. Photograph showing function


FIG. 102.- Case 7. Roentgenogram before operation, showing loss of substance

FIG. 103.- Case 7. Union five months after operation


Hospital. A bone graft 1¼ inches by three-sixteenths of an inch square was driven in a hole made in what remained of the base of the second phalanx and likewise introduced into the proximal end of the remaining shaft and head. Graft was self-retaining. Solid union occurred in two months, with union in the fractured third and metacarpal as well. July, 1923, a capsulotomy of the posterior capsular ligament of the metacarpophalangeal joint was performed, followed by normal function in the hand, there having been no injury to tendons. Returned to duty October, 1923.
CASE: 11. J. F., captain, Infantry. Gunshot wound, with compound comminuted fracture and loss of substance, right tibia, lower third, followed by osteomyelitis. Prior to admission,

FIG. 104.- Case 7. Photograph showing function

one bone graft had been done which was lost through infection. A large scar was removed, Walter Reed General Hospital, and a plastic closure of skin done. Operation, December 1,1922, Walter Reed General Hospital. Bone graft 10 ½ inches by three-fourths inch square taken from left tibia. The lower end was driven into the cancellous bone of the lower fragment. The upper end was inlaid into tile tipper fragment and the end of the graft fixed under the cortex, making it self-retaining. Patient was allowed to walk beginning the ninth month, wearing an ischial caliper. He was discharged December 2, 1923, with excellent bony union.
CASE 12. I. K. Gunshot wound causing compound comminuted fracture, 2½ inches loss of substance, lower third left tibia. Operation, April, 15, 1920, Walter Reed General Hospital. Sliding graft brought down from the upper fragment; bone held in place by kangaroo tendon. Periosteum sutured with catgut. Operation was followed by superficial infection, which healed. January 13, 1921. Patient fractured graft nine months after operation. Plaster cast applied and later a tibial caliper. Discharged January, 1922, wearing brace; clinically


FIG. 105.- Case 7. Another view showing function

FIG. 106.- Case 8. Loss of substance and atrophy present in humerus before graft


FIG. 107.- Case 8. Excellent bony union at end of six months

FIG. 108.- Case 8. Another view showing excellent bony union at end of six months


union was present. Returned for examination August 12, 1922, when there was solid bony union; more bone proliferation at the site of the fractured graft than anywhere else along the graft which bridged the loss of substance.
CASE 13. J. C. K. Gunshot wound, sustained in action, causing compound comminuted fracture, and 3 inches loss of substance tipper third left tibia, followed by chronic osteomyelitis. Operation, August 30, 1921, Walter Reed General Hospital. Bone graft, 7 inches by one-half inch, taken from inner surface right tibia, inlaid into both fragments and fixed with two kangaroo ligatures. Wound healed by primary intention. Roentgenogram November 17, 1921, when the original plaster was removed and replaced showing graft in excellent condition. Graft was fractured in plaster some three weeks later. Immobilization in a walking plaster cast or on an ischial caliper until May, 1923. Roentgenogram showed

FIG. 109.- Case 9. Lateral view showing comminuted fracture of patella and sepration of fragments

FIG. 110.- Case 9. Union present January, 1924

an attempt at union, but a pseudarthrosis occurred. May 15, 1923. A second bone graft was done; a graft 7 1/2 inches by 3/8 inch taken from the right tibia was inlaid into both fragments and outlaid along fractured graft, after freshening graft. The ends of the graft were fixed under cortical hone, making it self-retaining. Roentgenogram in August, 1923, three months after graft, showed union in the old fractured graft, and the new graft in excellent condition. Plaster was removed at this time and ischial caliper fitted. Patient discharged in January, 1924, with excellent bony union.
CASE 14. C. W. Gunshot wounds, causing compound comminuted fracture of tibia, right, and loss of substance lower third, with a chronic osteomyelitis. The left leg had been amp- tated through the middle third as the result of a gunshot wound. Operation, August 30, 1920, Walter Reed General Hospital. Bone graft taken from the upper third of the same tibia,


FIG. 111.-  Case 10 . Roentgenogram showing loss of substance and deformity


FIG.112.- Case 10. Two months after graft--deformity corrected with bony union


consisting of periosteum and osteum or cortical bone. A trough was made in both fragments, without opening into the medulla, with a notch cut in each end. The graft ends were sloped from above downward and the graft pushed in from the side, the ends fitting snugly in the notches in the cortical bone. The leg was immobilized in plaster before operation and the operation was done through a window in the cast. Solid bony union occurred, although the proliferation in the graft was slow. Patient discharged March, 1922.

CASE 15. H. W. Gunshot wound, with compound comminuted fracture and loss of substance and chronic osteomyelitis, lower third of both bones, right leg, 1½ inches above

FIG. 113.- Case 10. Showing function on completion

FIG. 114.- Case 10. Another view showing function on completion

the ankle joint. Patient bad had an unsuccessful " stepping " operation before admission. Two large scars were excised, one on the outer and the other on the inner surface of the leg before graft. Operation, May 19, 1922, Walter Reed General Hospital. Bone graft 41/2 inches by one-half inch taken from the inner surface of the left tibia. The lower end driven into a hole centrally placed in the lower fragment of the tibia to the depth of 134 inches, and inlaid into the upper fragment, the end of graft being fixed under the cortex, making itself-retaining. Excellent bony union occurred in the tibia, and in the fibula as well. Patient discharged wearing brace in May, 1923. He is now wearing no support.


FIG. 115.- Case 11. Roentgenogram showing loss of substance

FIG. 116.- Case 11. Roentgenogram three months after graft showing excellent condition of bone


FIG. 117.- Case 12. Roentgenogram showing loss of substance

FIG. 118.- Case 12. Roentgenogram three months after graft


FIG.119.- Case 12. Linear fracture ninth month

FIG. 120.– Case 12. Note absorption two months later


FIG. 121.- Case 12. Solid bony union 19 months after fracture

FIG. 122.- Case 13. Roentgenogram November 1921, fracture of first graft during fourth month and loss of substance bridged by graft


FIG. 123.- Case 13. Excellent union in old fracture in original graft and in new graft

FIG. 124.- Case 14. Roentgenogram of graft six weeks after operation


FIG. 125.- Case 14. One year after Figure 124, or 13½ months after operation, showing proliferation which had occurred in graft which had occurred in graft which bridged loss of substance


FIG. 126. - Case 15. No attempt at union in old fracture. Note proximity to ankle joint

FIG. 127.- Case 15. Lateral roentgenogram three months after graft


FIG. 128.– Case 15. Roentgenogram 11 months after graft. Outline of graft can barely be distinguished. Note union in fibula