Philadelphia Reflections

The musings of a physician who has served the community for over six decades

Related Topics

No topics are associated with this blog



Address reprints: Alvin H. Felman, M.D. Department of Radiology University Hospital 655 West 8th Street Jacksonville, FL 32209


I would like to thank the members of the Pennsylvania Radiological Society for inviting me to share in this marvelous celebration of Mary Fisher’s career in Radiology. Also, I want to express my thanks to this organization for recognizing the work that Mary has done over the years, and for consecrating, so to speak, what we, her former students, have known for so many years; namely that she is, indeed, entitled to be counted among the great radiologists of Philadelphia, of Pennsylvania, and indeed well beyond.

It is not my purpose to enlarge upon Mary’s career or her accomplishments. This singular honor that you have bestows says more than I or anyone could possibly say. Perhaps, however, I might add a few personal observations to this occasion form the standpoint of one who has known and worked with Mary.

She would arrive punctually each day in either 1953 Chevy or a large black Checker; either of which could defeat any Philadelphia weather system. Her day-long stint at her view box was interrupted almost ritualistically at five minutes to 12 each day for a trip to the hell vending machines where she purchased a package of peanut butter or cheeses crackers and soft drink which she consumed during the noon conference. At the end of each day, Mary disappeared into New Jersey where she cared for her children and husband George, but she could always be counted upon to be back at PGH for the evening conferences, ORP or other scheduled events.

Dr. George Fisher, known in some circles as the “Duke of Harrisburg”, is a fine physician and an author in his own right. According to Mike Huckman, a former resident, Mary was known as the Duchess of Altoona, a little apparently arising from some historic ties to British royalty.

I think it might be interesting to read part of a letter that I received from Dr. David Baker, Chairman of the Department of Radiology at Columbia. I wrote and asked Dave if he could uncover some past history of Mary Stuart, who was, I thought, a medical student at Columbia in the late 1940s and possibly also a radiology resident. He wrote me as follows:

“Dear Al:

Dashiell Hammett, Agatha Christine, Dorothy Sayers, and Dr. Arthur Conan Doyle collaborated and have finally found a picture of Dr. Fisher which I have made into a slide and I am sending to you. We had no record of Dr. Fisher as a resident but we were able to find Mary Stuart Blakely as a medical student who graduated in 1948. She interned at the Massachusetts General Hospital in 1949 and started her radiology residency here on October 1, 1949. She was appointed by Dr. Rose Golden. She wrote to Dr. Golden on July 30th, 1949 from Venice where she was spending the summer with her parents. The letter on July 30th noted that her address in Paris to the middle of August would be American Express Company, in Edinburgh until the end of August and in London through the first week in September. Sounds like a nice summer. A letter from Dr. Loeb to Dr. Golden on September 7 states that Dr. Blakely was the number one student in her class and was offered an internship at the Presbyterian Hospital but preferred erred acceptance at Massachusetts General where she went. While she was in medical school she won the Jane way Prize and was apparently an outstanding student. Her internship recommendations are equally glowing. Her record then became a little bit sparse.

The letter concludes with reference to the fact that Mary left Columbia for Bethesda and Washington and changed her name to Dr. Fisher in the interval.

I hope that this provides a little better insight into the background of your honored radiologist of the year.


The growing skeleton is at risk for epiphyseal and apophyseal growth plate injuries; these constitute approximately 15% of all the fractures in children. Both structural and functional differences exist between epiphyses and apophyses. Epiphyses, located at the ends of the long bones, are usually involved with articulations of joints. Their corresponding growth plates or physics are responsible for longitudinal bone growth and are composed primarily of cartilage and proliferation of cartilage cells. In addition to providing for growth, the epiphyseal physes function as resilient zones that resist forces of compression that occur in the running, jumping, and general play of growing children and adolescents.

Apophyses, in contrast to epiphyses, constitute promontories and excrescences of bone and function primarily as points of attachments for individual muscles or muscle groups. Their corresponding physes are less involved with growth but are organized to resist forces of tension that result from the contraction of muscles and the pull of ligaments. Apophyseal growth plates contain less proliferating cartilage and more fibro-collagen strands that bridge the physes, helping to prevent avulsion.


Injuries to epiphyseal growth plates are usually benign and heal without serious sequelae. However, certain types of trauma may cause damage to the proliferating cartilage cells with inhibition of growth, bony bridging across the growth plate, significant deformity, and functional impairment. The extent of these sequelae depends upon a variety of factors, chief among which are the extent of the original injury, location of the fracture, the future growth potential of the extremity, age of the patient, and ease and accuracy of the fracture reduction.

The Salter-Harris classification of fracture through the physes is useful in predicting the future growth arrest and other deformities. Ogden has proposed a more extensive scheme that takes into account several variations of the original Salter-Harris injuries and adds four additional types. (1) (Fig. 1) His classification illustrates serval rare injuries, occurring at birth or in the first two years that may have disabling consequences. Notable among these are types 1C and 2D where delayed growth arrest may occur as a result of damage to potential growth cartilage cells that are not as yet involved in the epiphyseal growth mechanism. Late epiphyses bar fusion become manifest when the epiphyseal growth plate expands to include the previously injured area.

Type 2 injuries are the most common overall and represent the most frequent type of physical separation in children over 10 years of age. Whereas these fractures usually heal without deformity, certain circumstances increase this risk. Damage to the growth plate, not present at the origin of the injury, may result from the act of reduction if the muscles a ligament are not relaxed. The undulation of the growth plate, especially of the distal femur, distal tibia, and proximal humerus make these areas especially vulnerable. In addition, the free metaphyseal fragments of Ogden type 2B may interfere with normal reduction.

Type 3 injuries occur most often when the physical growth plate is partially closed and invariably extend into the joint. Type 4 and 5 injuries are often associated with localized growth arrest and angular deformity despite the accurate reduction. Early definitive radiographic diagnosis and prognosis of this injury are often difficult or impossible leaving the clinician no choice but to treat expectantly. Sequential film studies may show relatively normal growth for serval years following which angular deformity may ensue.

With the development of the surgical technique for correcting partial growth plate closure, it is most imperative that these injuries be discovered and treated before the onset of serious deformity. (2, 3) Initial leg length measurements using accepted orthoroentgenographic technique should be obtained shortly after the initial injury. Thereafter, these measurements should be repeated at the at three to six-month intervals, depending upon the growth plate phase, in order to evaluate possible growth arrest. It should be remembered that the fractured bones, especially those of the lower extremities often heal with overgrowth and subsequent elongation. Hence, the failure of an injured bone to grow more rapidly than the comparison normal (bone) should be suspect for early epiphyseal growth arrest. (Fig.2)

When premature fusion of a growth plate occurs, surgical excision of the bony bar and insertion of one of a variety of materials (silastic, cranioplasty, etc.) is the accepted method of treatment. Radiological “mapping” of these abnormalities is necessary in order to evaluate the surgical approach and prognosis. Multidirectional tomography at 0.5 cm intervals in the frontal and lateral planes is used to map the size and location of the epiphyseal bridge (4). (Fig.3) Bright suggests the following criteria for surgical intervention: (1) partial growth arrest proven by tomography; (2) at least 50% of the normal remaining open growth plate; (3) at least two years of expected longitudinal growth; (4) adequate and good quality skin covering the lesion, and (5) free of drainage at least one year of previously infected. (2)

On occasion, early radiographic growth plate fusion may appear, only to resolve with a resumption of normal growth (Fig. 4) In these circumstances, one must weigh all clinical and radiological parameters before embarking on aggressive surgical treatment.


Apophyseal represent promontories of bone which generally serve as attachments for a muscle or muscle group. As much, their shape and configuration are usually determined by their loading requirements. Apophyseal growth plates of physes, while similar histologically to epiphyseal physes have significantly increased numbers of collagen fibers crossing the growth plates and considerably less proliferating cartilage cells. This morphologic alteration help withstand tension forces that result from contractions. Muscle attachments to apophyses are very strong by virtue of interosseous fibers that are extensions of the tendinous portion of the muscle. As a result, avulsion of portions of bone or fractures through apophyseal growth plates occur more regularly than torn tendons in the pediatric age group.

While injuries to apophyses may occur from a direct blow, most are caused by muscular effort. Often quite disabling, these injuries produce few important sequelae. On rare occasions, they may be confused with other, more significant lesion, i.e. infection, neoplasm, etc.

Avulsion injuries tend to be less acute in their onset; patients often present with a history of aching pain, usually of several weeks or months duration. In most instances, however, the patient experiences a sharp, intense pain associated with running, jumping or quick movements and changes of direction.

A Elbow:

Medial Epicondyle. This is the second center of the distal humerus to ossify; usually appearing between five and seven years. Fusion to the adjacent metaphysic does not occur until 17 or 18 years. As a result, the physics remains open during a protracted period of childhood and adolescence rending it susceptible to avulsion type injuries. (5, 6)

The medial Epicondyle serves as the attachment for flexors and pronators of the forearm and the ulnar collateral ligament. Serves valgus stress at the elbow is the mechanism that leads to avulsion of the medial Epicondyle. (Fig. 5) The medial Epicondyle also shares a physis with the trochlea which does not ossify until about 10 years of age. Hence, fractures extending through the medial Epicondyle may continue through the trochlea and into the elbow joint. This complication, though rare, must be recognized since external fixation is necessary to ensure proper healing and maintenance of function. (7)

The radiographic appearance of medial Epicondyle avulsion is usually characterized by considerable soft tissue swelling along the medial elbow and distal humerus and distal displacement of the ossific nucleus and loss of the normal parallel position in relation to the metaphysis, are additional findings. Not uncommonly, a “flake” of the adjacent metaphysic, pulled off along with the Epicondyle, serves as a confirmatory finding.

Extension of a medial Epicondyle avulsion fracture through the trochlea and into the elbow joint is difficult and often impossible to determine radiographically. The presence of a large metaphyseal fragment suggests extension through the trochlea but is most a definitive finding. Since the medial Epicondyle is often extra-articular, the absence of a joint effusion supports the diagnosis of isolated epicondylar avulsion; the presence of joint effusion, however, is of little help. Post reduction clinical and radiographic evaluations are probably as important as anything in determining the presence of an intraarticular fracture. Range of motion examination and radiographic appearance of the reduced fracture will usually suffice, (7) Arthrography should be reserved for those cases where definitive determination cannot be made.

The act of throwing, especially the use of a twisting movement when trying to throw a “curve-ball”, may also contribute to medial Epicondyle injuries. “Little League” elbow, a condition is seen in adolescent baseball pitchers and other players results in part from this action and the associated valgus stress applied to the elbow. (8) Repeated injury of the tendonous attachments to the medial Epicondyle may cause fragmentation of the ossification centers. (Fig. 6) On occasion, the medial Epicondyle may develop as several separate centers normally so that careful clinical and historical correlation is necessary if the diagnosis of “Little League” elbow is to be based on this finding alone. Comparison with the opposite elbow may be helpful.

The lateral condyle also may undergo change when subjected to the stress of throwing. Subchondral reabsorption, “cystic” lesions, and demineralization are among the recognized lateral condyle changes of “Little League” elbow. (Fig. 7) B Pelvis: Avulsion injuries usually resulting from various athletic endeavors represent the most common chrondro-osseous injuries about the pelvis. (10, 11) The anterior superior and anterior inferior iliac spines are points of attachments for the Sartorius and rectus femoris muscle respectively These are powerful flexors of the hip and may cause metaphyseal avulsions during running, licking and similar stressful exercises. While they often result from sudden muscular efforts and frequently evoke acute local pain, chronic aching in the area usually brings the patient to medical attention.

An apophyseal avulsion is usually recognized by an area of bony density within the adjacent soft tissue. The appearance varies depending upon the time interval between the injury and examination. The location of the injury also contributes to the roentgen appearance; ischial avulsion may be associated with exuberant callus formation. The roentgen appearance is usually typical, and when combined with the clinical history and symptomatology, is often conclusive. (Figs. 8, 9, 10,) Differentiation from a neoplastic process, while a theoretical dilemma, has not been a problem in our experience. Careful observation with repeated radiographs will usually clarify suspicion cases. Biopsy should be avoided unless there is a strong clinical suspicion of malignancy since healing avulsion fracture may be indistinguishable histological from malignant bone neoplasm.

C Knee:

Tibial Tuberosity. The quadriceps tendon inserts in a fan-like manner into the apophysis of the tibial tubercle and adjacent tibia. Acute fracture or avulsion of the tibial Tuberosity, or a portion thereof, may occur from direct trauma or from a sudden muscular contraction of the quadriceps mechanism. Classification of these injuries is based upon several features: 1) the degree of separation of the fracture from the adjacent metaphysis, 2) the accompanying soft tissue injury and possible functional impairment, and 3) extension into the tibial epiphysis with or without actual knee joint involvement. (Fig. 11)

The radiographic evaluation of these injuries is usually straightforward when the ossified structures are involved despite considerable normal developmental variations of the tibial tubercle. Accompanying soft tissue injury invariably occurs but the extent may not always be determined radiographically. Soft tissue edema in the region of the tubercle and loss of normal patellar tendon definition is usually evident but may escape detection if the film is not viewed with a bright light. The posterior edge of the normal patellar tendon should present a straight and well-defined interface with the infrapatellar fat pad. With soft tissue injury, this tendon may be widened and lose its distinct margin. Avulsed fragments of the tibial tubercle, (Fig. 12) The metaphyseal origins of these fragments is usually apparent since their borders are often irregular and a defect in adjacent tubercle metaphysis can frequently be identified. Dysplastic ossification within the patellar tendon may develop the following injury. (Fig. 13)

Attention must also be paid to the patellar-tubercle distance which frequently is increased when the tubercle is avulsed. This observation may be missed if the film is taken with the knee extended; flexion of the knee will accentuate this injury. Comparable films of the opposite extremity in the same degree of flexion may be necessary to confirm the diagnosis.

Type 3 injuries, with propagation into the tibial epiphysis and knee joint may disrupt the normal articular surface, (Fig. 14) In general, the radiographic detection of this injury is not usually difficult and when obscure, may be enhanced with conventional tomography. Replacement of an avulsed tibial tubercle to the normal anatomic position is necessary in order to restore complete function and avoid the complication of patella adults with possible secondary chondromalacia patellae.

Osgood-Schlatter’s “disease”, most commonly seen in adolescent males, was originally felt to be caused by aseptic necrosis of the tibial tubercle. In all probability, it results from partial traumatic fragmentation of the apophyseal ossification center, as a result of the strong tensile force applied to that area.

Although Osgood-Schlatter’s disease is usually a well-defined and easily diagnosed entity, other bone and/or soft tissue lesions may occur in this region. Arteriovenous malformation, osteomyelitis, benign and malignant tumors of bone and soft tissue have been reported. (13)

D Ankle:

The juvenile counterpart of the “Tillaux fracture” in adults is a Salter type III injury of the distal tibial physis. Closure of this growth plate begins in the mid-portion, progresses medially and then laterally. (14, 15) For a period of up to 12 to 18 months, the lateral physis may remain unfused rending it particularly vulnerable to a fracture. Extreme lateral rotation of the foot causes a stretching of the anterior tibiofibular ligaments leading to an avulsion type injury of the lateral epiphysis (Fig. 15.)

The radiographic diagnosis of juvenile Tillaux fracture is usually straightforward. Best seen in the mortise view, it produces a vertical fracture line through the epiphysis. (Fig, 16A) The epiphysis growth plate medial to the fracture line is fused; the lateral unfused portion may widen.

The plain film findings are diagnostic in most cases, but the use of computed tomography may be helpful both in the pre- and post- reduction evaluation (16). (Fig. 16B) When two or three part tri-plane fracture is suspected, the use of CT often is of great benefit. Complications and sequelae of these fractures are not common if the fragments are maintained in good alignment. The potential for premature growth arrest is minimal since they usually occur near the time of physiologic closure. Injury to the articular cartilage and failure to restore continuity of the joint surface carries a risk of future arthritic change.


Localized lesions and generalized conditions which weaken normal bone may result in fractures that occur with relatively minor trauma. Non-ossifying fibromas or benign cortical defects are common benign lesions usually discovered incidentally. On rare occasion, they present clinically as fractures with relatively minor trauma. (Fig. 17, 18) Because of the frequency of these lesions and the fact that they are so often discovered as films are taken for other reasons (trauma, pain, etc.), they may present problems in both diagnosis and management.

After the gross fracture has occurred, one has little choice but conservative treatment until healing occurs. The need to treat with current and packing before a fracture has occurred, and indeed after a healed fracture may present a dilemma for the patient and physician.

A high degree of individuality must be exercised in the approach to these conditions and must take into account the age of the child and the physical activity that is contemplated. While the latter can rarely be predicted, a more aggressive approach should naturally be considered in those patients who contemplate physically demanding sports.

The following roentgen criteria have been suggested as guidelines for the prophylactic surgical treatment of these lesions:

1. Non fibular location

2. Greater than 33mm long

3. Greater than 50% of bone width

4. Anticipation of physical contact or stressful Sports or other activity (17)

Fractures through simple cysts or other benign lesions of bone must be handled on an individual basis. Most bone cyst fractures will heal spontaneously, but there are no data to support the hypothesis that the cyst is more (or less) likely to heal as a result of a fracture. We have seen one case of malignant degeneration (osteosarcoma) in what was felt to be a six-year period.


A complete review of fracture complications is beyond the scope of this presentation. Nevertheless, I would like to illustrate a few unique situations where prompt recognition and proper treatment may prevent considerable morbidity and disability.

Pinckney, Currarino, and Kennedy have called attention to the damage of what might otherwise be considered a relatively benign injury, a stubbed toe. (18) They emphasize the risk of secondary osteomyelitis when Salter type I or II fracture of the epiphyseal growth plate of the distal phalanx of the great toe are associated with bleeding from the base of the nail bed or skin laceration in the immediate area. Prompt antibiotic therapy prevented this complication in two of their patients. We have seen a similar case in which osteomyelitis complicated a fracture of distal phalanx of the toe and a break in the overlying skin. (Fig. 19) The close relation of the skin to the periosteum, at the root of the nail with no intervening soft tissue, is the explanation given by these authors for the high incidence of complicating osteomyelitis.

The second complication of muscular skeletal injury is the presence of radiopaque foreign material at the site of injury. Dr. Fisher first suggested to me that all glass is radiopaque. After radiographing a large number and variety of glass objects, I was convicted of the truth of her statement (19). I have yet to see a proven case where the embedded glass was not visualized radiographically.


There are many historical accounts of the syndrome of child abuse but Dr. Fredric Silverman’s Rigler Lecture of 1972 should be read by everyone interested in this problem. (20) Dr. Silverman gives credit to Dr. Ambroise Tardieu, Professor of Legal Medicine at the University of Paris from 1861 to 1879, for first describing the feature of this syndrome. Tardieu outlined the demographic, social, psychiatric, and medical features that were, except for the roentgenographic findings similar in all respects to those that we recognize today. Caffey, in 1946, was probably the first t publish examples of the radiographic features of child abuse, but it wasn’t until 1953 when Silverman, one of his former pupils, corrected the clinical pattern of intentional trauma and abuse with the development of these characteristic fractures, (21, 22) The term, “The Battered Child” was coined by Kempe in 196, 100 years after Tardieu, 15 years after Caffey, and 10 years after Silverman called attention to this problem. (23)

Without trying to cover this subject in any substantial manner, I would like to stress a few salient features that may be of help in the diagnosis of this potentially fatal condition. Radiologists are usually alerted to the possibility of child abuse when skeletal surveys in films are requested. On occasion, however, one may recognize the “tell-tale” signs on film studies obtained for other reasons. I would like to illustrate several examples of child abuse that were recognized from chest studies requested for other reasons.

Rib fractures occur rarely in children from normal activity. When present, they should raise one’s suspicion of child abuse, especially when located in the posterior, paraspinal location, and when they appear to be in different stages of healing. (Fig. 20A) One is occasionally faced with the argument that rib fractures were caused by a parent attempting to “resuscitate” a chocking or apneic child. The presence of fracture of other bones, often the humerus, can be used as confirmatory evidence of child abuse. (Fig. 20B)

In recent years, cranial computed tomography has identified cases of child abuse that might have been overlooked previously. (24) Skulled fractures have been the traditional roentgen markers of cranial injury prior to the advent of this improved imaging technique (Fig. 21A, B). A wide variety of severe intracranial changes are now recognized in abused children, however, in the absence of skull fractures (Fig. 22.). These changes include subdual hematomas and effusions, cerebral contusion, and atrophy, etc. Cerebral contusions are most frequently seen and may appear as high density, isodensity, or low-density areas depending upon the relative amount of hemorrhage and edema. These injuries are thought to result from direct blunt trauma, accounting for extensive intracranial abnormality in the absence of skull fractures.

I want to again express my thanks to this society for allowing me to share in the honor of this day. Mary Fisher remains a symbol of excellence to which we, her former students and colleagues aspire. As long as we have teachers like her in our training programs, we will exhaust our supply of great radiologists. (Fig. 23)

In his farewell address to the graduating medical school class at the University of Pennsylvania in 1889, Sir William Osler read a poem about his revered mentor and great teacher, Dr. Palmer Howard. I can think of no better words to reflect my sentiments and those f my colleagues and former students of Mary Fisher. (Fig, 24)



1. Ogden J A: Skeletal growth mechanisms injury pattern. J Pediater Orthop 2:371 377, 1982.

2. Bright R W: Partial growth arrest: Identification, classification, and results of treatment. Orthop Trans 6:65, 1982.

3. Langenskod A: Surgical treatment of partical closure of the growth plate. J Pediatr Orthop 1:3 11, 1981.

4. Carlson W O, Wenger D R: A mapping method to prepare for surgical excision of a partial physical arrest. J Pediatr Orthop 4:232, 1984.

5. Chessare J W, et al: Injuries of medical Epicondyle. Ossification center of the humerus. Am J Roentgenol 129:49 55, 1977.

6. Silberstein M J, et al: Some vagaries of the medical Epicondyle. J Bone Joint Surg 6 A:524 528, 1981.

7. Cothay DM: Injury to the lower medial epiphysis of the humerus before the development of the ossific center, report of a case. J Bone Surg 49B: 766 767, 1967.

8. Brogdon BG, Crow N E: Little leaguer’s elbow. Am J Roentgenol 83:671 675, 1960.

9. Torg J S et al: The effect of competitive pitching on the shoulder and elbows of preadolescent baseball players. Pediatrics 49:267 271, 1972.

10. Fernbach S K, Wilkinson R H: Avulsion injuries of the pelvis and proximal femur. Am J Roentgenol 137:581, 1981.

11. Khoury M B et al: Bilateral avulsion fractures of the anterior superior iliac spines in sprinters. Skeletal Radiol 13:65 67, 1985.

12. Ogden J A, Southwick W O: Osgood-Schlatter’s disease and tibial Tuberosity development. Clin Orthop 116:180, 1976.

13. D’Ambrosia R D, MacDonald G L: Pitfalls in the diagnosis of Osgood-Schlatter’s disease. Clin Orthop 110:206 209 1975.

14. Kleiger B, Mankin H J: Fractures of lateral portion of the distal tibial epiphysis. J Bone Joint Surg 46A:25-32 1964.

15. MacNealy G A, et al: Injuries if the distal tibial epiphysis: systematic radiographic evaluation. Am J Roentgenol 138:683 689, 1982.

16. Cone R O, et al: Triplane fractures of the distal tibial epiphysis radiographic and CT studies. Radiology 153:763 767, 1984.

17. Arata M A, et al: Pathological fracture through non-ossifying fibromas: review of the Mayo Clinic experience. J Bone Joint Surg 63-A: 980 988, 1981.

18. Pinckney L E, et al: The stubbed great toe: a cause of occult compound fractures and infection. Radiology 138:375 377, 1981.

19. Felman A H, Fisher M S: Radiographic detection of glass in soft tissue. Radiology 92:524 526, 1966.

20. Silverman F N: Unrecognized trauma in infants, the battered child syndrome, and the syndrome of Ambroise Tardieu. Rigler lecture. Radiology 104: 337 353, 1972.

21. Caffey J: Multiple fractures in the long bones of infants suffering from chronic subdural hematoma. Am J Roentgenol 56:163 173, 1946.

22. Silverman F N: The roentgen manifestations of unrecognized skeletal trauma in infants. Am J Roentgenol 69:413 426, 1953.

23. Kempe C H, et al: The battered child syndrome. JAMA 181:17 24, 1962.

24. Tsai F Y, et al: Computed tomography in child abuse head trauma. CT, J Comp Tomo 4:277 286, 1980.

Originally published: Friday, October 06, 2017; most-recently modified: Wednesday, June 05, 2019