Background and Purpose: Hemipelvectomy is a life-changing treatment for pelvic malignancies. This study compared functional outcomes and quality of life of patients following internal or external hemipelvectomies.
Subjects: Ninety-seven patients who underwent tumor-related internal (n=39) or external (n=58) hemipelvectomy surgery between January 1, 1988, and December 31, 2004, participated in the study.
Methods: Using a descriptive retrospective cohort study design, functional status was evaluated with the Barthel Index at 3 time points. Quality-of-life parameters were evaluated at follow-up using the Linear Analog Self-Assessment tool (LASA).
Results: Data were collected on all 97 patients at rehabilitation consultation and hospital discharge. Follow-up data were obtained via survey on 44% of the original group at a median of 5.8 years (interquartile range [IRQ]=1.7–10.4) after surgery. Median total Barthel Index scores were similar between the internal and external hemipelvectomy groups at the initial physical medicine and rehabilitation assessment (10 [IQR=10–15] versus 10 [IQR=3.75–15]), at discharge (40 [IQR=30–65] versus 50 [IQR=35–66.25]), and at follow-up (92.5 [IQR=76.25–100] versus 92.5 [IQR=78.75–96.25]). Participants with external hemipelvectomies were less independent in bladder function and experienced greater pain severity at follow-up compared with those with internal hemipelvectomies. Overall quality-of-life parameters were similar between the groups.
Discussion and Conclusion: Despite significant differences in surgical procedures, immediate and long-term functional outcomes and quality-of-life parameters were similar among participants with internal and external hemipelvectomies.
Hemipelvectomy is a life-changing surgery that has been an acceptable form of treatment for pelvic malignancies for decades. Over the last 20 years, limb-salvage procedures, or internal hemipelvectomies, have emerged as an option for treatment.1 The oncologic treatments and outcomes of people requiring hemipelvectomy have been defined by several research teams; however, the rehabilitation outcomes after hemipelvectomy have not been well identified in the research, nor have residual bowel and bladder function been described.
The incidence and frequency of hemipelvectomy are unknown, but it is assumed that the majority of both internal and external hemipelvectomies are performed as primary treatments for sarcomas. The extent of the surgery is dependent on tumor location, size, and margins and involvement of vital neurovascular structures. Internal hemipelvectomy is the removal of the involved bone and adjacent muscles, with preservation of the lower extremity (Fig. 1). Internal hemipelvectomy is performed if a reasonably functioning extremity can be salvaged.2 External hemipelvectomy involves amputation of the lower extremity along with the hemipelvis (Fig. 2). Although surgery is the most common treatment for pelvic sarcoma, radiation therapy and chemotherapy may be used before or after surgery to shrink the tumor or kill remaining cancer cells.
Soft tissue and bone sarcomas are rare, with an incidence of approximately 6,000 to 7,000 soft tissue sarcomas and 2,500 bone sarcomas in the United States per year. These tumors represent less than 1% of all adult tumors.3 The tumor registry for the study facility reports an average of 11 new diagnoses per year of soft tissue sarcoma and an average of 12 new diagnoses per year of bone sarcoma between the years of 1988 and 2004.
Management of pelvic malignancies to optimize function is challenging due to the proximity of the tumor to the bladder, rectum, and neurovascular structures. Resection of these tumors requires complex surgeries and, in most cases, requires a collaborative surgical team. There is a 50% to 80% risk of complications following hemipelvectomy.1,4,5 In some situations, sacral nerve roots may be sacrificed when resecting the tumor, causing lower motor neurogenic bowel and bladder issues.6
Apffelstaedt and colleagues5,7 retrospectively reviewed 68 external hemipelvectomies performed between 1973 and 1994 and 32 internal hemipelvectomies performed between 1976 and 1994. These authors focused on surgical complications and mobility after these procedures. Their total mortality rates from the surgery were 6% for external hemipelvectomies and 9% for internal hemipelvectomies. They noted that postoperative complications occurred in 53% of the patients with external hemipelvectomies, with flap necrosis and wound infections being the most common. Internal hemipelvectomy postsurgical complications were similar (flap necrosis=12.5% and wound infection=47%). With respect to mobility, only 4.4% of the patients who had an external hemipelvectomy were able to use a prosthesis without external support; 81% used crutches, with or without prosthesis; 9% remained wheelchair-bound; and 6% were essentially bedridden. No comments were made regarding self-care independence, pain, bowel or bladder issues, or quality of life in either study.
In the 1970s to 1980s, internal hemipelvectomy was performed without reconstruction. Currently, there are reconstructive options for patients undergoing limb salvage. These options include arthrodesis, creation of a pseudarthrosis, allograft reconstruction, and insertion of a saddle prosthesis. Renard et al8 studied 15 patients following limb-saving reconstruction with a saddle prosthesis. The patients were typically on bed rest for 5 days postoperatively and began walking exercises 2 weeks postoperatively. Fifty percent of the patients were moderately functional at a 1-year follow-up. While this more recent publication better addresses function, it again primarily focuses on gait, with little information about other mobility issues and functional self care, bowel or bladder function, or quality of life.
Rehabilitation after hemipelvectomy requires a complex interdisciplinary approach to address the functional and emotional changes that occur. Further understanding of the functional outcomes for this population will assist the rehabilitation team in counseling patients, planning goals, providing treatment, and determining discharge needs and disposition.
The purposes of this study were: (1) to compare the functional status, including bowel and bladder management, of patients who had internal versus external hemipelvectomies at the time of initial physical medicine and rehabilitation (PM&R) assessment, at discharge from the acute hospital, and at follow-up and (2) to compare quality of life at follow-up between patients who have had internal and external hemipelvectomies.
The study population included all patients aged 18 years or older who underwent an internal or external hemipelvectomy secondary to tumor resection at a large midwestern hospital between 1988 and 2004 and had given consent for use of medical records for research. Patients who died during the initial hospitalization were excluded from analysis. Study participants were not offered remuneration.
Through the use of the facility's tumor registry, 103 patients were identified as having had pelvic tumors requiring hemipelvectomy. Five patients did not consent to use of their medical records for research, and one person's records were not sufficiently detailed to retrieve the data needed for the study. The remaining 97 patients in this study were divided into 2 groups: those with an internal hemipelvectomy (n=39) and those with an external hemipelvectomy (n=58). Table 1 shows the basic demographic data and characteristics of the study participants. A follow-up survey questionnaire was sent in the spring of 2005 to 71 patients who did not have a listed date of death in the medical record. Forty-three survey questionnaires were returned (response rate of 60%); 42 survey questionnaires were complete. Follow-up functional outcomes were obtained via patient survey using a modified Barthel Index (with permission granted by the Maryland State Medical Society to modify the tool to include layperson terminology). The Linear Analog Self-Assessment (LASA) was included in the survey to determine quality-of-life differences between the 2 groups.
Patient progress through the study is illustrated in Figure 3. Forty-three patients (24 in the internal hemipelvectomy group and 19 in the external hemipelvectomy group) returned the follow-up survey questionnaire. Median age at follow-up for the total group was 59 years (interquartile range [IRQ]=46–71) (57 years [IRQ=45–69] for the internal hemipelvectomy group and 60 years [IRQ=46–78] for the external hemipelvectomy group). Total years from surgery to survey for the entire group was 6 years (IRQ=2–10) (3 years [IRQ=1–10] for the internal hemipelvectomy group and 7 years [IRQ=3–10] for the external hemipelvectomy group).
Procedures and Measures
Retrospective patient data were obtained by chart review, which included demographic information, cancer diagnosis, and functional outcomes during acute hospitalization. Functional abilities of patients at the time of the initial PM&R assessment and at discharge from the acute hospital setting were documented by the rehabilitation team. Functional abilities were determined from previously entered clinical data using the Barthel Index.9,10 The Barthel Index is a measure of functional capacity (Appendix 1). The Barthel Index is a reliable, valid, and widely used tool that assesses 8 self-care activities, including bowel and bladder management, and 2 mobility activities with an ordinal scale.9,10 The total of the itemized scores ranges from 0 (total dependence) to 100 (complete independence).
Quality of life after hemipelvectomy was measured using the LASA tool, which is a public domain tool that has been validated as a general measure of global quality of life in numerous settings.1,4,6,11–14 Locke et al14 described the LASA as a valid short instrument to assess quality of life in patients with cancer. They found the LASA items correlate with similar, well-validated multiple-item scales such as Functional Assessment of Cancer Therapy–Brain, the Symptom Distress Scale, and the Profile of Mood States (P<.001 in each case). The LASA is a 10-point linear analog scale that assesses overall quality of life; mental, physical, emotional, and spiritual well-being; severity and frequency of pain; social activity; and level of support (Appendix 2).
The primary outcome variables of the study were Barthel Index scores (total and subcategory scores), pain, length of hospitalization, and mobility. Medians and IRQs were determined for dependent variables. The Wilcoxon rank sum test was used for between-group (internal hemipelvectomy group versus external hemipelvectomy group) comparisons for pain index, length of time from surgery to mobility, functional outcomes, and total Barthel Index scores at each time point. The Wilcoxon rank sum test also was used for between-group (internal hemipelvectomy group versus external hemipelvectomy group, men versus women) comparisons for total Barthel Index scores at each time point. Frequency distributions were obtained for each functional assessment item of the Barthel Index. When comparing 2 nominal variables such as surgeries performed in 5-year time frames and discharge disposition of each group at discharge and at follow-up, the chi-square test or Fisher exact test was used when appropriate. The chi-square test also was used to compare percentage of total independence between groups (internal hemipelvectomy group versus external hemipelvectomy group).
Descriptive statistics (medians and IRQs) were used to analyze quality-of-life (LASA) individual items. The Wilcoxon rank sum test was used for between-group comparisons.
Participant Mobility in the Postsurgical Period
The percentages of internal versus external hemipelvectomy surgical procedures remained consistent through the study time frame (P=.99) (Tab. 2). Participants who received external hemipelvectomies had longer acute hospitalizations (median=20 days [IQR=12.75–29]) compared with participants who received internal hemipelvectomies (median=14 days [IQR=10–19]) (P=.02). Prescribed bed rest, ordered by the surgeon, occurred more often (P=.0107) in the external hemipelvectomy group (29%) than in the internal hemipelvectomy group (8%). Only 4% of the participants in both groups developed flap necrosis or wound infections. The amount of time elapsed from surgery to the first time the patient stood was longer for participants with external hemipelvectomies (median=4 days [IQR=2–7]) compared with participants with internal hemipelvectomies (median=2 days [IQR=1–6.5]) (P=.04). No difference was found in days postsurgery to ambulation between the 2 groups (median=6 days [IQR=3–11] for the external hemipelvectomy group and 4 days [IQR=2–7] for the internal hemipelvectomy group).
Cohort Analysis of Functional Outcomes
Total Barthel index scores at the PM&R assessment, at discharge from the acute hospital setting, and at follow-up were compared between the 2 groups. No differences in functional abilities between the groups were observed at the 3 different time points (Tab. 3). Interestingly, at the PM&R assessment, there was no difference in functional abilities between men and women. However, at discharge, women were more independent in function than men (P=.005). Functional abilities at follow-up were similar between women and men.
Both groups were dependent in functional abilities at the PM&R assessment; however, by follow-up, both groups were nearly fully independent. Table 4 presents the percentages of participants who were independent (scored as maximum points) in Barthel Index subcategories at discharge and at follow-up. At discharge, a greater percentage of participants with external hemipelvectomies were independent in transfers compared with those with internal hemipelvectomies (P<.01). At follow-up, a greater percentage of participants with internal hemipelvectomies were independent in bladder function compared with those with external hemipelvectomies.
Disposition after hemipelvectomy was similar between the 2 groups at discharge (P=.90) and at follow-up (P=.49). Half of the participants in both groups returned home from the acute setting, and the other half transferred to a skilled nursing facility, an acute rehabilitation facility, or a swing bed. Upon follow-up, all of the participants with external hemipelvectomies were residing at home (26% were living alone, 74% lived with family). Ninety-two percent of the participants with internal hemipelvectomies resided at home (4% were living alone, 88% lived with family), and the remaining 8% were in a skilled nursing facility or an assisted-living setting.
Forty-two participants (72%) with external hemipelvectomies were discharged with a sitting orthosis. Of the 19 participants who submitted follow-up information, only 5 (26%) continued to use the sitting orthosis. Of interest, only 2 of the 19 participants (11%) were using a prosthesis for gait on a regular basis. Despite this, 79% of the participants with external hemipelvectomies reported being completely independent with mobility, without requiring use of a wheelchair.
Quality of Life
On LASA subcategories, no differences were noted between groups for any parameter except pain severity. Participants with external hemipelvectomies experienced a higher level of pain severity relative to those with internal hemipelvectomies. Most of the pain reported by the external hemipelvectomy group was phantom pain, not ongoing incisional or surgical pain. Participants with external hemipelvectomies experienced more severe pain and bladder dysfunction than those with internal hemipelvectomies; however, overall quality-of-life scores were similar between the 2 groups (Tab. 5).
This retrospective study found no difference in function at 3 different time frames between 2 significantly different surgical procedures: internal hemipelvectomy (limb sparing) and external hemipelvectomy (amputation). Our findings showed that few participants were independent in functional abilities at discharge but that the majority of participants in both groups were totally independent with transfers, mobility, feeding, bathing, grooming, and toileting at follow-up (an average of 6 years later).
Overall quality of life was similar between the groups at follow-up. The only differences that were identified between the 2 groups were that participants with external hemipelvectomies had greater ability to transfer at discharge but increased pain severity and bladder dysfunction at follow-up compared with participants with internal hemipelvectomies.
Surgical complications can impair mobility and function. Apffelstaedt and colleagues5,7 and Higinbotham et al15 reported 47% to 75% wound complications, respectively, including hematoma, seroma, infection, and skin necrosis. The most common limitation to mobility after surgery in the current study was prescribed bed rest, rather than flap necrosis or wound infection. Only 4% of the participants had flap necrosis or wound infections. Prescribed bed rest has been indicated by surgeons to protect the incision and flap, thus lowering the risk for postoperative complications.8
Renard et al8 studied 15 patients with internal hemipelvectomies who underwent saddle prosthesis surgery. Median time to ambulation was 18 days for patients with primary tumors and 8 days for those with secondary tumors. In our study, median time from surgery to ambulation was not different between the external hemipelvectomy group (6 days) and internal hemipelvectomy group (4 days); however, these time frames were much shorter than those found in the study by Renard and colleagues.
Median length of stay postoperatively was 6 weeks in the study by Renard et al.8 In the studies by Apffelstaedt and colleagues,5,7 the average length of stay after external hemipelvectomy was 39 days for curative resection and 24 days for palliative resection, and the average length of stay after internal hemipelvectomy was 35 days for curative resection and 25 days for palliative resection. In the current study, participants with external hemipelvectomies had longer acute hospitalization stays compared with those with internal hemipelvectomies, which is commonly attributed to complications after surgery. Overall, our study showed shorter lengths of stay postoperatively (median of 14 days for the internal hemipelvectomy group and 20 days for the external hemipelvectomy group) in comparison with previous studies. This finding is reflective of a shorter time frame from surgery to ambulation, despite prescribed bed rest and fewer postsurgical wound complications (4%).
The functional ability of people after hemipelvectomy has been discussed only briefly in recent reports. Pring et al2 studied patients with chondrosarcoma who underwent either internal or external hemipelvectomies. All of the patients available for follow-up who underwent internal hemipelvectomies were able to walk; however, this was not quantified. In the study by Renard et al,8 a 1-year follow-up of 11 patients indicated that all were ambulatory but that the distances they were able to ambulate were variable between household ambulators and community ambulators. In the studies by Apffelstaedt and colleagues,5,7 34% of the patients with internal hemipelvectomies were ambulating independently at the time of discharge and 59% were ambulating with crutches. The mobility status of people who underwent external hemipelvectomies was reported to be 4% independent with a prosthesis and 81% ambulating with a gait aid; however, the time frame was not identified in the reports.
In the current study, independence in ambulation (mobility) was quantified, using the Barthel Index, as ambulating more than 45.72 m (50 yd), with or without a gait aid but without assistance of another person. A greater percentage of participants with external hemipelvectomies were independent in transfers at discharge. This result may have been due to the need for spica-type bracing after an internal hemipelvectomy, thus requiring more assistance with transfers. There was no difference between the 2 groups regarding independence in mobility at discharge and at follow-up. The majority (>78%) of both groups were independent in mobility at follow-up. Although there was no difference between the groups in ability to negotiate stairs, the results of the current study emphasize the difficulty of negotiating stairs after such procedures, as only 52% of the participants with external hemipelvectomies and 71% of the participants with internal hemipelvectomies were independent at follow-up.
Discharge disposition is based largely on a person's functional abilities and support at home to provide assistance needed. This study did not intend to predict discharge disposition via functional outcomes. However, at the time of discharge, the majority of the participants in both groups were not independent in transfers, mobility, stairs, bathing, dressing, or toileting. Thus, more than half of each group required discharge to a skilled nursing facility, an acute rehabilitation setting, or a swing bed to continue therapy to improve their functional abilities, which is reflective of the follow-up data. Functional outcomes assist not only in discharge disposition planning but also in determining equipment, home modifications, and home health care assistance for patients returning home.
Similar to the study by Apffelstaedt et al,7 there was low use of prostheses in the external hemipelvectomy group in the current study. Despite their availability for this population, prostheses require a large amount of energy for functional ambulation. Thus, discussions regarding prosthetic fitting should be emphasized more for those who will have the strength and endurance for use. Many times the prosthesis is prescribed more as a tool for standing and for cosmetic reasons than for gait. The treatment team must be clear about its goals for the patient. In fact, mobility may be much easier and safer for some patients without a prosthesis. In our study, sitting orthoses were commonly molded for participants with external hemipelvectomies to distribute seating pressure evenly. Long-term use of these devices was low among the study participants. Currently, sitting orthoses are rarely prescribed during the acute phase of rehabilitation secondary to existing drains at the time of discharge and the availability of more supportive and adaptable seating devices for wheelchairs.
Bladder function after hemipelvectomy had not been studied until the current study. A greater percentage of participants who underwent an external hemipelvectomy were dependent on others for bladder function, especially at follow-up. This finding may be attributed to the surgical need to sacrifice sacral nerve roots for tumor resection. This information is extremely important for patients to be aware of prior to surgery, as many do not anticipate the loss of bladder function associated with their amputation. Continence is psychosocially important to patients; however, lower motor neurogenic bladders can be difficult to manage. The results are a flaccid detrusor and external sphincter that cause the person to be incontinent and unable to empty the bladder completely, leading to increased risk of urinary tract infections. Bladder management methods are individualized, with the goals of protecting renal function and upper and lower urologic structures and maintaining continence.
Although there were no differences between the groups in bowel function, both groups remained impaired at follow-up. Sacrifice of sacral nerve roots can result in lower neurogenic bowel. Loss of anal tone and lack of external sphincter contraction can cause fecal retention and oozing of stool. Again, it is imperative that patients be informed of these potential changes and that they be managed with an individualized bowel program.
Quality of life after hemipelvectomy has scarcely been reviewed. Refaat et al16 noted similarity in the quality of life of patients who had amputation and those who had limb salvage for treatment of high-grade sarcomas of the lower extremity. Seventy-two percent (n=66) of the amputation group and 74% (n=342) of the limb-sparing group were satisfied with their current life status. Results of the current study mirror those of Refaat and colleagues, despite that fact that participants with external hemipelvectomies experienced greater severity of pain (phantom pain) relative to participants who underwent internal hemipelvectomies. Health care providers must be aware of the effect of pain on quality of life, community re-entry, and community involvement. Neuropathic pain, such as phantom pain, can be a challenge for health care providers, but its management is essential for the patient to continue to be active within his or her family and community.17
It is important for the patient and health care team to know that the likelihood of being functionally independent at the time of discharge is low. With time, practice, and possibly additional therapy, many people with either internal or external hemipelvectomies can become independent in most functional activities. The results of our study will enable the health care team to help the patient understand potential outcomes and discharge disposition following surgery. For example, people undergoing an external hemipelvectomy should be counseled to anticipate a longer length of stay and potential risk for bladder dysfunction and phantom pain.
Limitations and Future Directions
The Barthel Index was modified for use in the survey; however, the modified version of the tool was not validated. The survey also created possible inconsistencies in patient self-reported function and quality of life. One of the 43 survey questionnaires was not completed. The decision to use the LASA, a quality-of-life instrument for which reliability and discriminative validity have yet to be firmly established, is a methodological weakness that could influence the interpretation of results. Chart review is associated with possible discrepancies due to missing data. Of the 97 charts reviewed, 13 charts were missing 5 or fewer data points and 2 charts were missing between 6 and 10 data points. The results from this single-center study can be generalized only with the following cautions: the limited number of participants with a variation in health status, the limited number of surgeons performing the surgery, and multiple rehabilitation team members.
The comparison between internal and external hemipelvectomy in this study, generally speaking, was a comparison of participants who underwent amputation and those who underwent limb-sparing procedures. However, there can be considerable variation in function within the limb-sparing group alone. This variation is dependent on whether a partial hemipelvectomy or a total hemipelvectomy was performed; some surgeries may result in nerve compromise as well. Further studies on the internal hemipelvectomy group alone may provide a wealth of information to health care providers.
This retrospective study used a mixed source of data, retrospective chart review, and current status survey. A prospective study of this population would rectify data logistics as well as provide clearer follow-up needs of the patient.
This is the most comprehensive study thus far describing the functional outcomes and quality of life of patients after tumor-related hemipelvectomy. Function and quality of life in patients with hemipelvectomy are comparable despite significantly different procedures of limb salvage versus amputation. The only differences that were identified between the internal and external hemipelvectomy groups in this study were that participants with external hemipelvectomies had greater ability to transfer at discharge but increased pain severity and bladder dysfunction at follow-up compared with participants with internal hemipelvectomies. However, it is important to note the potential loss of bowel and bladder function related to both surgeries and the effect on the patient. Additional research with prospective longitudinal data is needed to identify functional progress among people with hemipelvectomy. Details regarding bowel and bladder dysfunction and management warrant further evaluation.
Ms Beck, Ms Einertson, Dr Winemiller, and Dr DePompolo provided concept/idea/research design. Ms Beck, Ms Einertson, and Dr Winemiller provided writing. Ms Beck and Ms Einertson provided data collection, project management, fund procurement, and subjects. Dr Winemiller provided data analysis. Dr Hoppe and Dr Sim provided consultation (including review of manuscript before submission).
Permission was granted to modify and reprint the Barthel Index using layperson terminology from: Mahoney FI, Barthel DW. Functional evaluation: the Barthel Index. Md State Med J. 1965;14:61–65. Used with permission.
The Mayo Clinic Institutional Review Board approved this study protocol.
A poster presentation of this work was given at the 67th Annual Assembly of the American Academy of Physical Medicine and Rehabilitation; November 9–12, 2006; Honolulu, Hawaii.
- Received June 26, 2007.
- Accepted April 7, 2008.
- American Physical Therapy Association