Survival And Prosthetic Fitting Of Children Amputated For Malignancy

C.B. Taft

Report of a study conducted by the staff of New York University Child Prosthetic Studies under a special grant from the Children's Bureau, Department of Health, Education and Welfare.

Special acknowledgment is made of the contributions of T.Ann Gorton, Nancy Munson, and Hector W. Kay in the conduct of the study and preparation of the manuscript.


In the management of children who have had an extremity removed because of a malignant tumor, a difference of opinion sometimes arises as to whether or not the child should be fitted with an artificial limb. It is generally agreed that the restored function and cosmesis afforded by prosthetic fitting are of practical and psychological value to both the child and parents. However, because of the uncertain prognosis and the time and cost involved in prosthetic restoration, it has been contended by some that early fitting is unwarranted.

One reason some agencies delay provision of limbs is the limited availability of data relating to survival for periods of less than five years after amputation. Most statistics are based on five-and/or ten-year survival rates which reflect lower percentages of longevity. For example, the American Cancer Society reports that 17.3% of a sample of 595 patients with a history of bone tumor survived five to ten years after surgical treatment, and 14.2% of a second sample of 442 survived ten years or longer 1.

Data from recent studies indicate that a significant percentage of amputees with malignancy survive one to five years postoperatively and wear prostheses successfully for one year and longer. Aitken 2 reported that 17 children who died had worn their limbs an average of 1.2 years; 11 children who were alive and under care at the date of his survey had an average wear period of 2.8 years; and 14 children who were still alive when lost to follow-up, had worn their prostheses an average of 3.2 years. In an unpublished study at the University of Illinois, Lambert 3 found that the average period of wear for 42 children who were fitted with prostheses was 3.5 years. Twenty children who died had worn their prostheses an average of 1.08 years; 22 children still living and under care had an average wear period of six years. Freed and Charette 4 reported similar data on a study of 12 patients, aged 13 to 60 years, who were amputated because of malignancy. Nine surviving subjects had been followed from 1.8 years to 6.3 years postoperatively, and 3 who died had lived 2.9 to 3.8 years postoperatively. The majority of these patients were full-time users of their prostheses. In a study of 10 patients (19 to 57 years of age) with hemipelvectomies, Watkins 5 reported that postoperative survival ranged from 1 to 20 years and that seven of these patients were successful limb wearers. None of these authors advocates a delay in prosthetic restoration. It was believed that a limb should be fitted as soon as the stump is healed, provided that there is no evidence of metastasis. The consensus is that the expenditure of time and money involved is vindicated by the favorable rehabilitation results.

In order to obtain additional data on the postoperative survival of children amputated for malignancy and on the prosthetic fitting of these children, a survey of 23 cooperating juvenile amputee clinics was conducted by the Child Prosthetic Studies of New York University at the request of the Subcommittee on Child Prosthetics Problems, Committee on Prosthetics Research and Development, National Academy of Sciences-National Research Council. The primary purpose of the survey was to gether data which would provide a statistically meaningful basis for the prosthetic management of amputees with a history of malignant tumor.

Method Of Study

Questionnaires (Fig. 1 ) were sent to 23 child amputee clinics in 16 states, the District of Columbia, and the Canadian provinces of Ontario and Quebec. Each clinic was asked to examine its records and to furnish information on each patient who had been amputated for a tumor prior to the age of 21 years. Data were collected over a period of four months, and the composite data were then analyzed on two bases: (1) total sample (T.S.) the entire group of 350 patients reported; and (2) complete data sample (C.D.S.), limited to the 278 patients on whom complete information had been provided. The patients in both samples were categorized into three groups: Group A - those still living and under care; Group B - those alive when lost to follow up; Group C - those who had died.


Total Sample

Twenty-two Juvenile Amputee Clinics (See Appendix I) reported data on a total of 350 children on whom amputations had been performed because of malignancy, during the period from 1943 to 1965. There were 199 males (56.9%) and 151 females (43.1%) in the series. Table 1 presents the distribution of the 350 children according to survival status (Groups A, B, C).

The study would not have been possible without the generous cooperation of these participating clinics. Their assistance is gratefully acknowledged.

It will be noted that 219 (62.6%) were reported alive, either at the time of the survey or when lost to follow-up. Of these 219 patients, 120 (54.8%) were males and 99 (45.2%) were females. Of the 127 patients who were deceased, 76 (59.8%) were males and 51 (40.2%) were females.

Distribution of the 350 cases by histologic diagnosis, sex, and extremity amputated is presented in Table 2 . Twenty-six different diagnoses were reported, including the non-specific diagnoses of "tumor" and "sarcoma". The five specific types of malignancy listed in the table accounted for 283 cases or 80.9% of the total. For the remaining 67 cases (19.1%) there were 21 diagnoses, none of which included more than 6 cases. These cases are grouped under "Miscellaneous".

Distribution of the 350 cases according to the level of amputation is presented in Table 3 . As may be seen, 301 (86.0%) had amputations of the lower extremity and 47 (11.4%) had amputations of the upper extremity, a ratio of approximately 6-1/2 to 1. This finding is almost identical to the 7:1 lower to upper extremity ratio in the 47 cases reported by Aitken 2, who emphasized this large proportion of lower extremity involvement and remarked that "the provision of some mechanism in order that this large group may ambulate without the use of crutches seems almost mandatory". It may be noted that this situation is the reverse of that found in congenital limb deficiencies, in which defects of the upper extremity significally outnumber those of the lower extremity.

Complete Data Sample (C.D.S.)

Because of incomplete information, 72 cases could not be included in the analysis of certain important aspects of the study, e.g., age at amputation, postoperative longevity, time lapse between amputation and prosthetic fitting, or the length of prosthetic wear. Of the 278 cases available for complete analysis, 158 (56.8%) were males and 120 (43.2%) were females. The distribution of these 278 children in the three status categories (Groups A, B, C) is presented in Table 4 . As may be seen, 173 children (62.2%) were reported alive either at the time of the survey or when lost to follow-up (Groups A and B). Of the 173 children in these two groups, 93 (53.8%) were males and 80 (46.2%) were females.

Distribution of the 278 cases by histologic diagnosis, sex, and extremity amputated is provided in Table 5 . The five specific diagnoses listed in the table comprised 224 cases or 80.6% of the total. In the sample of the 278 children, 239 (85.0%) had lower extremity amputations and 38 (13.7%) had upper extremity amputations, a ratio of approximately 6 1/2 to 1. Comparing Table 4 and Table 5 (C.D.S.) with Table 1 and Table 2 (T.S.), it will be seen that the distribution of cases according to the various categories is remarkably similar.

Age At Amputation

Table 6 presents the distribution of the 278 children by age at amputation. The age range was from 2 weeks to 21 years and 11 months, with an average of 10 years and 11 months. The average age at amputation was: Group A - 10 years and 5 months, Group B - 11 years and 1 month, and Group C - 13 years and 11 months.

As may be seen in Table 6 , a total of 84 children (30.2%) were amputated between the ages of 0 and 10 years, whereas 187 children (67.2%) were amputated between the ages of 10 and 20 years. The largest number of amputations occurred in the 14-16 year age period, involving 52 children or 18.8% of the sample.

Longevity after Amputation

Data on the postoperative longevity of these children are of primary interest. These data are summarized in Table 7 , Table 8 , Table 9 and Table 10 and graphically presented in Fig. 2 . Postoperative longevity of the total sample of 278 children ranged from two months to twenty years and one month (Table 7 ). Particularly significant are the findings that 214 children (77.0%) - more than three-quarters of the total number - were alive more than a year after amputation. Of these, 149 children had lived from one to five years; and 65 children had lived five years or longer after surgery. These data, of course, tend to underestimate the true situation in that some additional children of the 49 lost to follow-up (Group B) would presumably live to swell these survival figures; as would some of the 124 children still under care (Group A).

A total of 105 children had died (Group C). The postoperative survival of these children ranged from two months to five years and three months, with an average of 1.5 years. It may be observed that 70 of these children or two-thirds of the total lived more than one year after amputation. Postoperative longevity of the 173 children still under care or lost to follow-up ranged from two months to twenty years and one month, with an average of 4.4 years in Group A and 4.9 years in Group B.

Table 8 presents the data on survival in relation to age at amputation. As previously noted, 84 children or 30.2% of the total were amputated between the ages of 0 and 10 years and 187 children (67.2%) were amputated between the ages of 10 and 20 years. Referring to Table 8, it will be seen that 16 of the total of 105 deaths (15.2%) were in the 0-10 years group and 85 (81.0%) were in the 10-20 years group. Another way of expressing these data is that: of 84 children amputated between the ages of 0 and 10 years, 19.07. had died, as contrasted to 45.4% of the 10-20 year group (85 of 194 children), thus indicating a more favorable prognosis for children in the 0-10 year age period. There were no significant differences between these two groups, however, in the duration of postoperative survival of those who died.

It is of considerable interest that the largest number of deaths and the highest mortality rate for any four-year period occurred with children who were amputated between the ages of 12 and 16 years. As may be seen, these 52 deaths comprised 54.7% of the 95 children amputated at these ages and accounted for 49.5% of the total of 105 deaths. For comparison, the mortality rate in the 8-12 year age group was 34.4% and the mortality rate in the 16-20 age group was 35.7%. This significantly higher mortality rate in the 12-16 year age group suggests that the prognosis for survival is poorest for children of these ages. It should be noted that for those who died there were no significant differences between the three age groups cited in length of survival after amputation. Type of Malignancy in Relation to Mortality and Longevity

Because of the high incidence of osteogenic sarcoma and the associated mortality rate, it seemed desirable to evaluate its relationship to the above findings. Comparative data on mortality and postoperative longevity as related to age at amputation and diagnosis are presented in Table 9 , Table 10 , and Table 11 .

Patients with osteogenic sarcoma constituted about 437. of the 84 amputations in the 0-10 year age group, with a mortality rate of 33%. In the 10-20 year age group the percentage of osteogenic sarcoma cases rose to 65%, with a mortality rate of 55%. Likewise, the overall mortality rate for other types of tumors rose from 8% in the 0-10 year age group to 28% in the 10-20 year age group. It may be observed that the overall mortality rate for other types of tumors increased markedly in the 8-12 year age group to a peak of 32% and remained at this level until the end of the 12-16 year age group, when it declined materially. On the other hand, the mortality rate for osteogenic sarcoma remained fairly constant until the 12-16 year period, whereupon it rose sharply to a peak of 64% or almost two-thirds of the 67 cases of osteosarcoma. It remained at this level until the end of this period, when it also declined.

It is obvious that the peak in the overall mortality rate occurring in the 12-16 year age group was caused primarily by the concomitant increase in the mortality rate for osteogenic sarcoma. The reasons for the marked increase in the mortality rate of osteogenic sarcoma at this time are subject to conjecture. However, it is to be noted that this period coincides with the associated marked growth spurt and other changes which occur in connection with puberty and adolescence.

Table 10 presents the data concerning survival after amputation as related to diagnosis. Table 11 represents these data in terms of 1-plus years and 5-plus years post-amputation survival, together with data taken from the "Cancer Prognosis Manual" of the American Cancer Society 1. Five-year survival rates after surgical treatment of malignant bone lesions, including osteogenic sarcoma, fibrosarcoma, Ewing's sarcoma, and chondrosarcoma are shown.

Referring to the 5-year survival rates, it should be emphasized that our data deal only with patients under 21 years of age, whereas the data in the "Cancer Prognosis Manual" are based on patients of all ages. According to our survey, 16.0% of the 163 cases of osteogenic sarcoma reported and 23.4% of the total sample of 278 survived for at least five years after amputation.

It will be noted that the 5-year rate of 16.0% for osteogenic sarcoma is very close to the corresponding rate of 18.4% from the "Cancer Prognosis Manual". However, the 5-year rate of 23.4% from our total sample is about one-third greater than the 17.3% of the "Cancer Prognosis Manual". This difference is apparently due to higher survival rates in our data in the fibrosarcoma and "Miscellaneous" categories. The numbers of cases of Ewing's sarcoma, chondrosarcoma and synovial sarcoma are too small to permit significant comparisons.

Prosthetic-Fitting and Wear

Prostheses were fitted to 243 children, or 87.4% of the total sample of 278 on whom complete data were available. Of the remainder, nine children (3.2% of the sample) had limbs prescribed but had not worn them; seven because of the rapid course of the disease and two because they were amputated just prior to the survey. Prostheses were not prescribed for 26 children, 9.4% of the sample, for the reasons presented in Table 12 .

For the children who were fitted with a prosthesis, the time lapse between amputation and fitting ranged from two weeks to twelve years and five months. The majority of patients (55.1%) were fitted within six months after surgery (Table 13 ).

The data obtained on the known periods of prosthetic wear for the 243 children who were fitted provide the most meaningful basis for a conclusion concerning the desirability of early prosthetic fitting (Table 14 ).

The length of prosthetic wear ranged from one week to 13 years and 10 months. As indicated in the table, 164 children (67.5%) or more than two-thirds of the total number fitted had worn their prostheses for periods of one to more than 13 years.

For the group of 81 children who died while under care (Group C), the period of wear ranged from 1 month to 5 years, with an average period of 1.2 years. Forty-six (56.8%) of these children had worn their limbs for a minimum of one year, while 16 (19.7%) had worn them for 2 to 5 years.

The length of prosthetic wear by the 162 children fitted in Groups A and B ranged from 1 week to 13 years, with an average wear period of 4.4 years. Of this total of 162 children, 118 (72.8%) wore their prostheses for at least one year, while 87 (53.7%) wore their limbs 2 years or longer. The number of children achieving these goals would continue to increase because of the open end nature of the study.

Agency Policies

Twenty-one clinics in 15 states, the District of Columbia, and the Canadian provinces of Ontario and Quebec provided information on the policies of their Crippled Children's agencies regarding time lapse between amputation and provision of a prosthesis to be purchased by the agency. The overwhelming majority stated that there are no agency requirements in this respect, the decision concerning lapse of time before fitting being left to the discretion of the physician. Agency restrictions were reported in only three states as follows: no local or pulmonary metastasis during postoperative periods of three months (one state) and six months (two states). The practice followed by the great majority of the reporting clinics was to fit the prosthesis as soon as possible after amputation, if the stump was suitable for fitting and chest X-rays revealed no metastasis.

Summary And Conclusions

A survey of child amputee clinics in the United States and Canada was conducted to obtain data on the post-amputation survival and prosthetic history of children who had been amputated because of malignant bone tumor. Twenty-two clinics reported on a total of 350 children who had been amputated for malignancy between 1943 to 1965.

Of 278 cases on whom complete data were available for analysis, 214 (77.0%) lived at least one year after amputation. Of these, 149 lived for 1 to 5 years postoperatively, and 65 lived 5 years or longer following amputation. Of 243 children who had been fitted with prostheses, 164 (67.5%) wore their prostheses for periods of 1 to 13 years. The incidental finding of a marked increase in the mortality rate of osteogenic sarcoma during the period of puberty and adolescence suggests the need for further study of this observation.

These findings provide convincing evidence that amputation because of malignant bone tumor does not constitute an "a priori" contraindication to early prosthetic fitting. More positively stated, it would appear that if there is no evidence of metastasis and the stump is suitable for fitting, a limb should be fitted at the earliest possible date. To quote Aitken (2): "Adherence to such a philosophy would provide many children with an extended period of active, useful function."

Appendix I: Reporting Amputee Clinics

  1. Area Child Amputee Center, Grand Rapids, Michigan

  2. Institute for Physical Medicine and Rehabilitation, New York, New York

  3. Newington Hospital for Crippled Children, Newington, Connecticut

  4. University of Illinois Amputee Clinic, Chicago, Illinois

  5. Birmingham Child Amputee Clinic, Birmingham, Alabama

  6. Georgia Juvenile Amputee Clinic, Atlanta, Georgia

  7. Children's Rehabilitation Center, Buffalo, New York

  8. Child Amputee Prosthetics Project, UCLA, Los Angeles, California

  9. University of Oklahoma Hospital, Oklahoma City, Oklahoma

  10. Kernan Hospital Amputation Clinic, Baltimore, Maryland

  11. Child Amputee and Congenital Deficiency Clinic, Seattle, Washington

  12. Child Amputee Clinic, Orlando, Florida

  13. Home for Crippled Children, Pittsburgh, Pennsylvania

  14. Child Amputee Clinic, Memphis, Tennessee

  15. Child Amputee Clinic, Elizabethtown, Pennsylvania

  16. Juvenile Amputee Clinic, New Orleans, Louisiana

  17. Sunnyview Rehabilitation Center, Schenectady, New York

  18. Kessler Institute for Rehabilitation, West Orange, New Jersey

  19. Juvenile Amputee Clinic, Washington, D. C.

  20. Ontario Crippled Children's Centre, Toronto, Ontario, Canada

  21. Rehabilitation Institute of Montreal, Montreal, Canada

  22. Shriners Hospital, Springfield, Massachusetts

C.B. Taft is the Assistant Project Director and S. Fishman, Project Director at the New York University Child Prosthetic Studies


1. James, Arthur G., Cancer Prognosis Manual, American Cancer Society, Inc., New York, 1961, p. 73. 
2. Aitken, George T., Prosthetic Fitting Following Amputation for Bone Tumor, A Preliminary Report, Inter-Clinic Information Bulletin, 3:1-2, 10 (February-March) 1964. 
3. Lambert, Claude N., Personal Communication. 
4. Freed, Murray M., and Charette, Edmond E., Rehabilitation After Amputation of the Lower Extremity for Malignancy, Arch. Phys. Med. and Rehab., 45:564-570 (November) 1964. 
5. Watkins, A.L., Rehabilitation After Hemipelvectomy, J.A.M.A. 181:793-794, (September 1) 1962. 
6. Loon, H.E., The Past and Present Medical Significance of Hip Disarticulation, Artificial Limbs 4:4-19 (Autumn) 1957. 
7. Gordon-Taylor, G., The Incomputable Factor in Cancer Prognosis, Brit. Med. J. 1:455-462 (February 21) 1959. 
8. Watkins, A.L., Prosthetic Rehabilitation After Hemipelvectomy, Ortho. and Prosthetic Appliance J. 17:173-174 (June) 1963. 
9. Phelan, John T., Grace, James T. and Moore, George E., Hemipelvectomy for the Management of Soft Tissue Tumors of the Lower Extremity, Amer. J. of Surg. 107:604-608 (April) 1964.
10. Shands, Alfred R., "Handbook of Orthopaedic Surgery", St. Louis, the C. V. Mosby Company, 5th Edition, 1957. 
11. ______, Subcommittee of the Statistics Committee, Manual of Tumor Nomenclature and Coding, corrected edition, New York, American Cancer Society, 1953. 
12. ______, Committee on Prosthetics-Orthotic Education of the National Academy of Sciences-National Research Council, Amputee Census Revised October 1964. 
13. Coventry, Mark B. and Dahlin, David C, Osteogenic Sarcoma; A Critical Analysis of 430 Cases, J. Bone and Joint Surg. 39-A:741-758 (July) 1957.