Quality of life after allogeneic hematopoietic cell transplantation

Quality of life (QOL) is a dynamic, multifaceted concept related to physical, cognitive, emotional, and social functioning and well-being. Issues related to QOL are routinely cited by cancer survivors as among their greatest concerns.¹  QOL is an especially important consideration in the counseling, implementation, and posttreatment management of arduous treatments for life-threatening conditions, such as allogeneic hematopoietic cell transplantation (HCT). Although potentially lifesaving, allogeneic HCT carries an attendant risk of significant acute complications, late effects including chronic graft-versus-host disease (GVHD), organ toxicity, osteoporosis, infections, cataracts, secondary cancers, and infertility, as well as decrements in QOL.^2-5  A growing literature has described the impact of HCT on QOL, but is characterized by heterogeneity in study design, patient population, comparator groups, assessment instruments, time points examined, and conclusions reached. A cohesive overview is needed to integrate these studies and describe the clinical and research implications of findings.

The aim of the current review is to synthesize and critically evaluate the current literature on QOL after allogeneic HCT in adults. We start by addressing issues in the assessment of QOL in HCT. We also examine the concept of clinically significant change in QOL. We then summarize studies in a way we believe to be most relevant to clinical practice, including longitudinal change in QOL in allogeneic HCT and comparisons of QOL in allogeneic HCT with autologous HCT, standard-dose chemotherapy, healthy matched comparators, and population norms. An examination of GVHD, reduced-intensity conditioning regimens (RICs), and other predictors of QOL follows. Patient-reported benefits of HCT and behavioral interventions to maintain or improve QOL are then described. We conclude with a discussion and recommendations to researchers and clinicians.

A Medline search was conducted using MeSH terms “quality of life” or “health related quality of life” and “bone marrow transplantation” or “hematopoietic stem cell transplantation.” Further searches were conducted based on related references identified in Medline and examination of references cited in selected articles. Criteria for selected articles were (1) HCT for hematologic malignancies, (2) sample of adults at time of assessment, (3) inclusion of patients who underwent allogeneic HCT, and (4) use of at least one quantitative multi-item measure of QOL. Excluded studies were those exclusively focused on pediatric populations, only autologous HCT, available only in non-English languages, or otherwise not relevant.

Although there is recognition of the importance of standardized assessment of QOL across clinical trials,⁶  there is currently no consensus regarding which measure should be used. Instead, the HCT literature encompasses a variety of QOL measures, including broad measures of QOL in healthy and patient populations, measures of cancer-specific QOL, HCT-specific QOL, and GVHD-specific side effects (Table 1). General measures of QOL have the advantage of applicability for both patients who underwent HCT and comparison groups, but may be less sensitive to side effects of transplantation, including acute and chronic GVHD, than HCT-specific measures.^7,8  Thus, investigators may wish to supplement a general QOL measure with a HCT- or GVHD-specific measure to adequately capture QOL.

There is also significant heterogeneity in the content of QOL measures, as seen in Table 1. Even measures assessing similar domains of QOL may differ in content. In a comparison of the EORTC QLQ-C30 and the FACT-BMT, Kopp et al⁹  noted that whereas total scores of the 2 measures were highly correlated, subscales assessing emotional functioning demonstrated a low correlation. The authors suggested that these subscales may not be directly comparable.⁹  In addition, systematic differences may result from the way QOL measures are administered, with interview administration generally resulting in better reported QOL than paper-and-pencil administration.¹⁰  These differences may be due to patients' tendency to not report less severe symptoms during interviews.¹¹  It should be noted that in the current review, no systematic differences in findings were evident by measure or method of administration. Nevertheless, these issues should be considered when selecting measures of QOL in patients who underwent HCT.

Although research typically evaluates the statistical significance of QOL differences within or between patients and comparison groups, equally important is the clinical significance of findings. Clinical significance refers to the implication, clinical utility, or practical importance of the QOL data.^12-14  Methods used to examine clinical significance include anchor-based and distribution-based methods. Anchor-based methods use an independent standard for interpretation of QOL study results, such as cutoff scores. For example, differences of 5, 10, and 20 points on the EORTC QLQ-C30 indicate small, medium, and large effects, respectively.¹⁵  In clinical practice, anchor-based methods are useful to examine change in individual patients. In research, anchor-based methods can be used to interpret mean differences or report the percentage of patients reporting clinically significant change or impairment.^8,16-18  Distribution-based methods interpret results in terms of magnitude of mean difference and variability in a sample of subjects, such as effect size. Distribution-based methods can easily be applied to any continuous measure in a sample reporting means and standard deviations (SDs) or standard errors, although the results in a given sample cannot be applied to a new sample. This method is widely used in research^18-20  but is not suited to clinical applications.

The current review uses both anchor- and distribution-based methods to describe clinically significant QOL outcomes when possible. In regard to anchor-based methods, we report study results in terms of proportion of those survivors of HCT who achieve a certain clinically important beneficial or deleterious outcome (eg, return to work). We use distribution-based methods to report study results in terms of effect size. This effect size has been calculated in the following manner: in longitudinal studies, the difference in QOL scores between the study end point and baseline are divided by baseline SD for each domain. In between-subjects comparisons, differences in mean QOL scores between the HCT sample and comparison group are divided by the SD for the comparison group for each QOL domain. Effect sizes can be interpreted in accordance with the Cohen guidelines, in which .2, .5, and .8 SD correspond to small, medium, and large differences, respectively.²¹ 

Studies of QOL in HCT patients are summarized in Table 2. A fundamental issue in longitudinal studies of patients who underwent HCT is attrition due to mortality and study withdrawal, which ranges from 29% to 65% by 1 year after HCT.^2,16,22-25  Importantly, several studies suggest that patients with better QOL are more likely to have complete data,^2,24-27  which raises the possibility of bias when reporting results only for patients who have remained in the study. Although investigators may be tempted to use traditional intent-to-treat analyses to examine this issue, we would argue that imputing QOL data for patients who have died makes results difficult to interpret. Existing studies have attempted to produce unbiased estimates of posttransplantation functioning in the following ways: Jacobs et al²⁸  have used statistical modeling to estimate and control for attrition by time interaction effects on objective cognitive impairment; Lee et al²²  examined dichotomous outcomes (eg, return to work, good vs poor QOL) and reported percentages of patients who died, survived with good outcomes, survived with poor outcomes, or were missing data. The probability of a surviving patient reporting a good outcome (ie, conditional analysis) and the probability of a patient surviving and reporting a good outcome (ie, unconditional analysis) were reported.²²  We describe the results of conditional and unconditional analyses separately.

Among patients who survive transplantation, physical functioning rapidly declines immediately after transplantation, reaching a nadir at 30 to 100 days.^{2,23,24,27,29,30}  Physical functioning then begins to improve, with 2 studies^2,24  finding that physical functioning plateaus in the year after transplantation and another study²⁶  finding continued improvement over 4 years. Regarding the magnitude of improvement, 3 studies reported that mean levels of posttransplantation physical functioning were higher by .1 to .2 SD compared with pretransplantation levels,^2,23,24  2 reported that they were lower by .8 to 1.3 SD,^29,30  and 2 did not report SD.^16,17  Syrjala et al³¹  reported some fluctuation in physical functioning over time, with 25% of patients reporting major physical limitations at baseline; 44%, at 90 days; 12%, at 1 year; 22%, at 3 years; and 18%, at 5 years. In general, longitudinal studies suggest that survivors can expect long-term physical functioning similar to or worse than that before transplantation.

Transplantation appears to take the greatest emotional toll on survivors early in the process. Emotional functioning for survivors is most compromised before transplantation and immediately after,^23,24  with significant improvements seen as early as hospital discharge²³  to 100 days.^24,29,30  Some data suggest that emotional functioning remains relatively stable after this initial improvement,^23,29  although other data suggest that it continues to improve in the 2²⁴  to 4²⁶  years after transplantation. In general, survivors can expect average improvements in emotional functioning of .3 to .4 SD from before to after transplantation.^23,24,29,30 

Regarding social functioning, patients who undergo transplantation fare relatively well. Data are conflicting regarding short-term social functioning in survivors, with one study finding improvement by .4 SD from baseline to 90 days after transplantation,²⁹  another finding deterioration by .4 SD from baseline to 100 days,³⁰  and a third finding no change.²³  At 1 year, survivors reported similar or better levels of social functioning compared with pretransplantation baseline,^16,23,29  with significant improvements at 3 years¹⁷  and continued improvements in years 1 through 4.²⁶  Long-term improvements in social functioning were of the magnitude of .3 SD.²⁶  By 2 years after transplantation, 84% of survivors reported enjoying socializing with friends and family, up from 52% at 6 months and 77% at 1 year.²² 

Role functioning shows an immediate decline after transplantation followed by gradual improvement over time. From baseline to 90-100 days, decrements of .6 SD were observed in role functioning,³⁰ .7 SD in work functioning,²⁹  and .8 SD in home management.²⁹  One year after transplantation, 2 studies suggest that role functioning has returned to baseline levels,^16,17  whereas another²⁹  found that work functioning had improved to .4 SD and home management had improved to .2 SD above pretransplantation levels. By 1 year, 59% to 69% of survivors have returned to work, school, or homemaking.^16,22,26  Data suggest continued improvement in role functioning of .4 SD from year 1 to year 4.²⁶  These data are corroborated by increases in return to work, school, and homemaking during this time. Bush et al reported that 80% of survivors had returned to work or school at 2 years, 80% at 3 years, and 74% at 4 years.²⁶  More than 84% of survivors had returned to work or school by 5 years or more.^2,5  Thus, it appears that the majority of survivors of allogeneic transplantation resume their roles at home and in the community after transplantation.

Survivors also report good overall QOL, particularly as time from transplantation increases. Overall QOL follows a similar pattern to role functioning, with deficits immediately after transplantation,^23,24  and return to baseline levels at day 100.^23,24,30  After day 100, overall QOL may stabilize or continue to improve, with patient ratings similar to or better than baseline at 6 months,^25,32  1 year,^22-25,32,33  2 years,^22,24  and 3 years.^16,25  Long-term improvements in overall QOL relative to baseline were of a magnitude of .3,²³  .6,^24,33  and 1.1³²  SD. Bush et al²⁶  found that 1, 2, 3, and 4 years after transplantation, 73%, 76%, 81%, and 80% of survivors rated their overall QOL as good to excellent, respectively. By 2 years after transplantation, 71% of survivors reported that they had recovered from their transplantation, up from 41% at 6 months and 66% at 1 year.²² 

Unconditional estimates of post-HCT QOL have resulted in mixed findings. Using intent-to-treat analyses, 2 studies have reported longitudinal improvement in post-HCT QOL.^26,27  Bush et al²⁶  found that 1 to 4 years after transplantation, unconditional mean estimates of QOL were high and were generally comparable with conditional estimates. Domains of QOL improved or stayed the same over time, with social functioning, role functioning, and physical functioning showing significant improvements of .5, .4, and .3 SD, respectively. Altmaier et al²⁷  imputed the worst possible QOL rating for patients who had died and half of the maximum rating for patients who were too ill or otherwise did not complete an assessment. They found no statistically significant differences in mental or physical QOL at 1 and 3 years after transplantation compared with pretransplantation baseline. HCT symptomatology was increased at 100 days after transplantation but returned to baseline levels by 1 year. In contrast, Lee et al²²  reported that the probability of surviving and returning to work, school, or homemaking was 26% at 1 year and did not appreciably change at 2 years.²²  The probability of surviving and reporting health to be very good or excellent was 18% at 1 year and 13% at 2 years.²²  As missing data were included in the denominator but not the numerator, these estimates may be low. Nevertheless, estimates of survival with good and poor outcomes highlights the variability of post-HCT QOL and provides information that is easily understandable to patients. Continued research and use of innovative methodologies are needed to produce estimates of QOL after HCT independent of study completion.

Comparisons of QOL between patients undergoing allogeneic and autologous HCT are complicated by multiple confounders, including age, pretransplantation comorbidities, and higher rates of relapse in autologous patients. Although several observational studies have compared QOL in these groups, few have controlled for confounding variables. Before transplantation, allogeneic patients experience higher QOL, including global QOL, role functioning, physical functioning, and cognitive functioning, than autologous patients.^16,17  After transplantation, one of the largest studies to date found no differences by type of transplantation at 90 days and 1, 3, and 5 years, controlling for a variety of medical and sociodemographic variables.²  These findings are supported by a second study that found no differences in QOL between groups in the first 100 days after transplantation.³⁴  In contrast, other data suggest that allogeneic patients recover more slowly from transplantation.^17,22  Allogeneic patients showed a more dramatic decline in QOL, which returned to baseline levels by 4 to 8 months after transplantation, whereas autologous patients recovered to baseline levels of functioning 2 to 4 months after transplantation.¹⁷  At 6 months after transplantation, allogeneic patients were more likely to have seen a physician in the past month, more likely to still be on medication, and less likely to have resumed work, school, or homemaking than autologous patients.²²  Data are similarly conflicting regarding long-term QOL. Three studies have reported that allogeneic patients report worse QOL 1 year after transplantation and beyond, including poorer overall QOL,^25,35  physical functioning,^35-37  role functioning,^35,36  and social functioning³⁶  than autologous patients. Four other studies have reported nonsignificant differences in QOL 1 year or more after transplantation.^16,22,38-40  Finally, one study reported that allogeneic patients experienced statistically significant better role functioning and clinically (but not statistically) significant better cognitive functioning, social functioning, and overall QOL than autologous patients at 3 to 5 years after transplantation.¹⁷  Available data indicated that the magnitude of difference in physical functioning was .5 to .7 SD worse in allogeneic patients.^37,38  Taken together, the majority of available data suggest that patients undergoing allogeneic transplantation report QOL that is similar to or worse than patients undergoing autologous transplantation, but these findings may be confounded by older age, greater comorbidities, and increased rates of disease progression in autologous patients.

Similar to the comparisons described in “Comparisons with autologous HCT,” comparisons between allogeneic transplantation and standard-dose chemotherapy may be confounded by systematic differences in patients who undergo allogeneic transplantation versus chemotherapy. Conflicting findings characterize this literature. Three studies have reported worse QOL after HCT, including physical,^35,36  role,^36,41  cognitive,⁴¹  emotional,⁴¹  and social functioning^36,41  as well as overall QOL.^35,36  Four studies reported no statistically significant differences between HCT and standard-dose chemotherapy.^16,17,37,39  One study reported better physical and psychological functioning after HCT but worse overall QOL.⁴²  Effect sizes for physical functioning were .4 SD better⁴²  and .2 to .4 SD worse^39,41  in HCT. For psychological functioning, effect sizes were .2 SD better⁴²  and .2 to .4 SD worse^39,41  in HCT. HCT displayed .3 SD worse role functioning,⁴¹  .4 SD worse cognitive functioning,⁴¹  .4 SD worse social functioning,⁴¹  and .1, .2, and .5 SD worse overall QOL.^39,41,42 

Interesting differences in findings emerged by study methodology. All 3 studies recruiting only patients in remission reported worse QOL in the group that underwent transplantation^35,39,41 ; in 2 of the 3 studies these differences reached statistical significance.^35,41  Thus, comparing only patients in remission, patients treated with HCT exhibit worse QOL than patients treated with chemotherapy. Four of the 5 remaining studies that did not specify remission as an eligibility criterion^17,37,42,43  reported better QOL in the group that underwent transplantation; in one study⁴²  these differences reached statistical significance. Thus, QOL appears to be better after HCT than chemotherapy when patients with and without relapse are included. This pattern of findings suggests that differential rates of relapse in HCT and chemotherapy patients may influence QOL findings, with higher rates of relapse in the chemotherapy group potentially associated with worse QOL in that group. However, as none of the studies reported rates of relapse by group, further research comparing HCT with standard-dose chemotherapy with specific attention to rates of relapse is now needed.

Studies have consistently demonstrated statistically significant impairments in patients who underwent HCT relative to both healthy controls and population normative data. Comparisons with adults without cancer at assessment points ranging from 5 to 10 years after transplantation have found decrements in patients who underwent HCT in physical,⁴⁴  social,⁴⁴  psychological, and emotional functioning, as well as overall QOL.^16,20,44,45  Effect sizes were .7 SD for general health,²⁰  .3 to .5 SD for physical functioning,^20,44  .4 SD for role functioning,⁴⁴  .1 to .4 SD for psychological functioning,^20,44  .4 to .6 SD for social functioning,^20,44  .5 SD for cognitive functioning,⁴⁴  and .2 SD for overall QOL.⁴⁴  Survivors of transplantation view their overall health as worse than a typical person their age but they also report greater personal growth.²⁰  In addition, an average of 42 months after transplantation, 25% of patients who underwent transplantation reported significant medical problems, whereas 15% to 25% reported emotional distress, low self-esteem, and low life satisfaction.⁴⁶  Thus, data suggest differences of moderate size between survivors of HCT and adults without cancer on average, with significant percentages of patients reporting impairment.

In comparison with population normative data at time points ranging from 6 months to 18 years after HCT, statistically significant decrements in QOL are again seen in patients.^40,47-52  Ten years after transplantation, the magnitude of these differences was .7 SD for general health, .3 SD for physical functioning, and .5 for physical role functioning.⁴⁷  Interestingly, although the majority of studies suggest that patients who underwent HCT report worse QOL than population norms regardless of time since transplantation, one study suggests that this only holds true for patients fewer than 3 years out from transplantation; patients who were greater than 3 years after HCT had QOL scores that were comparable or better than population norms.⁴⁹  Examining only patients less than 3 years after HCT, decrements of .8 SD were observed in general health, .9 SD in physical functioning, 1.3 SD in role functioning-physical, .5 SD in role functioning-emotional, .6 SD for social functioning, and .2 SD in mental health.⁴⁹  Nevertheless, the self-reported QOL in the survivors represented in these studies is largely positive, with 67% to 80% reporting overall QOL as “good to excellent,” and 74% reporting QOL as “same or better” than pre-HCT levels.^48,51,52  In total, the studies represented here indicate statistically and clinically significant decrements in QOL in survivors of HCT compared with both healthy controls and population normative data. However, the majority of HCT survivors represented here report their QOL as “good to excellent,” and this proportion provides an avenue for conveying the clinical meaning of these results from clinicians to patients.

GVHD shows a robust, negative relationship with QOL. Of 7 studies investigating the relationship between acute^8,47  and chronic^{8,16,29,47,53-55}  GVHD and QOL, 6 have reported a significant, negative relationship.^{8,29,47,53-55}  Only one study¹⁶  has found no relationship between chronic GVHD and QOL, but it may have been underpowered. Both acute and chronic GVHD have been shown to be associated with worse physical functioning,^29,55  role functioning,^53-55  social functioning,^53,55  mental health,^47,54  general health,^47,54  and overall QOL.⁸  Calculated effect sizes indicate moderate to large differences of .5 SD for physical functioning, .7 SD for role functioning, and .7 SD for social functioning.⁵⁵  Only 60% of patients with chronic GVHD were able to work.⁵⁵  Syrjala et al also reported that chronic GVHD predicts impaired physical recovery at 1 year after HCT.²⁹  There is less consistency regarding resolved chronic GVHD, with one study suggesting that it is associated with worse QOL⁴⁷  and another suggesting that is not.⁵⁴  Further studies are needed to examine lingering effects of resolved chronic GVHD on QOL.

The literature examining the impact of RIC on QOL is limited and bears further exploration. Overall, studies suggest that QOL after RIC is good⁵⁶  and comparable with that seen with myeloablative conditioning²⁴  as well as autologous HCT.⁵⁷  Patients receiving RIC appear to have an immediate QOL advantage over autologous patients, but this advantage may be reversed 3 to 6 months after transplantation.⁵⁷  However, these comparisons are confounded by patient factors that influence the decision to use RIC regimens. More work needs to be done to elucidate the impact of RIC on post-HCT QOL.

A maturing body of research has examined the role of other sociodemographic, patient, disease, and treatment factors in determining post-HCT QOL. Poorer QOL is associated with greater degree of symptoms, lower educational level, older age, shorter period after HCT, female sex, sexual impotence, advanced disease at time of transplantation, presence of chronic GVHD, worse pretransplantation level of functioning and impairment, greater interpersonal conflict, and reduced level of social support.^{29,40,48,50,54,58-61}  Predictors of post-HCT QOL are an important area for future research, as identification of early predictors could contribute significantly to better informed consent for treatment.

Several authors have reported on the transformative nature of HCT, and potential positive impact on psychological and interpersonal growth, or “posttraumatic” growth.^20,62-66  Specifically, patient-reported benefits include an enhanced appreciation for life, different priorities, love and appreciation for family and friends, and greater religious or spiritual beliefs.^20,64,65  Reported benefits are unrelated to physical functioning and other QOL indices.^20,66  These data suggest that patients are often able to reinterpret the adversity of HCT into a meaningful life narrative despite reduced QOL.

There is widespread interest in behavioral interventions to improve quality of life after HCT. Research has examined the effects of exercise and psychosocial interventions in patients who underwent allogeneic transplantation. Among randomized controlled trials of hospitalized patients who underwent HCT, supervised exercise resulted in better patient-reported physical well-being at discharge⁶⁷  and significantly smaller decline in physician-rated performance status⁶⁷  and maintenance of muscle strength⁶⁸  at 100 days compared with usual care. Effects were stronger for patients who were less fit before transplantation and those receiving RIC.⁶⁷  Exercise was generally well-tolerated during the hospitalization period, with 68% of the treatment group reporting exercise 5 times a week or more while hospitalized.⁶⁷  Further, inpatients reported several benefits of exercise, including improved strength and energy, alleviation of boredom, increased endurance, maintenance of flexibility, and emotional distraction.⁶⁷  Among outpatient survivors of HCT, supervised treadmill walking^69,70  and home-based aerobic exercise⁷¹  are associated with decreased fatigue,^70,71  increased physical well-being,⁷¹  and increased aerobic fitness.^69-71  One study⁷⁰  noted that improvements in fatigue were maintained 1 year after the intervention, indicating that the beneficial effects of an exercise program may be sustained. Magnitude of effects were .4 SD for physical functioning and .01 SD for psychological functioning.⁷¹ 

Psychosocial interventions in HCT have been examined in 2 randomized controlled trials.^72,73  Neither examined QOL as an outcome but were instead designed to test the effects of stress management and coping skills training on pain, nausea, and emesis compared with usual care and a time and attention control. In both trials, patients in the stress management and coping skills groups reported reduced pain.^72,73  Coping skills training did not appear to enhance the effects of stress management on pain. Additional studies are needed examining QOL as an outcome.

In summary, behavioral interventions show promise to maintain or improve quality of life after allogeneic HCT, consistent with a larger body of evidence regarding the benefits of exercise and stress management in cancer patients.^74,75  The HCT literature has been plagued by small sample sizes and high levels of attrition due to death and complications. Larger randomized controlled trials examining the effects of exercise and stress management on QOL are needed.⁶  In the meantime, clinicians should consider recommending moderate-intensity aerobic exercise several times a week for patients who are able to engage in such activity and who may benefit from it. Clinicians should also consider a psychosocial referral for training in stress management for patients experiencing pain.

The current review synthesized existing literature examining QOL in adults treated with allogeneic HCT and reported the clinical significance of these outcomes. These findings demonstrate the statistically and clinically significant adverse impact of hematopoietic stem cell transplantation on QOL. Longitudinal studies support early decrement in QOL after allogeneic HCT,²³  but largely report improvement to baseline levels thereafter. The literature reviewed here suggests impairments in QOL and differences in the trajectory of recovery after HCT in allogeneic HCT compared with both autologous HCT and chemotherapy; however, several important studies do not support this conclusion, and confounding variables in these comparisons have been inadequately examined. Cross-sectional studies suggest impairment in QOL compared with healthy controls or population normative data. The literature reported here supports acute and chronic GVHD as threats to QOL, and uncertainty remains regarding the ongoing impact of resolved chronic GVHD. Although the literature on RIC is limited, it does not suggest a clear advantage of RIC over myeloablative conditioning or autologous HCT. Behavioral interventions have demonstrated efficacy in improving QOL, and bear further exploration.

The moderate deficits in self-reported QOL described by existing literature are somewhat surprising in light of the many objective impairments experienced by patients who underwent HCT. Findings may result from averaging QOL across subgroups of patients doing well and others doing poorly. One study³³  noted greater variability in post-HCT QOL compared with before transplantation, a pattern that is also evident in other longitudinal studies and between-group comparisons.^32,42,44  Response shift may also contribute to discrepancies between self-reported QOL and objective impairments. A major challenge in the design and interpretation of studies of QOL, response shift is defined as a recalibration of internal standards of measurement, a reconceptualization of the meaning of items, and a reprioritization of values.⁷⁶  That is, patients may evaluate QOL differently after significant declines in QOL, with QOL that was once considered poor now considered acceptable or good. Emerging literature has demonstrated the presence of response shift in other medical conditions such as HIV/AIDS and prostate cancer.^76-78  It is likely that response shift plays a similar or even more important role in patients undergoing HCT as this arduous treatment and its associated complications may be an even more potent catalyst for change in appraisal of QOL.⁷⁷ 

Several major challenges remain for investigators. First, efforts to reduce measure-based heterogeneity through the use of standardized instruments would facilitate comparisons between trials and greatly accelerate efforts to study post-HCT QOL.^6,79  Second, attrition needs to be examined as a major threat to the validity of longitudinal studies, through use of both conditional and unconditional analyses. Next, a greater understanding of longitudinal recovery after HCT requires a systematic examination of response shift, which has yet to be examined in HCT. In addition, we would recommend that investigators communicate their study results both in terms of statistical significance and in terms of clinical relevance; this will serve to enhance the interpretation of these results by researchers, clinicians, and patients. Further work remains to be done in relation to the long-term impact of resolved chronic GVHD, the impact of RIC regimens on QOL, and other predictors of impaired QOL after HCT.

Several challenges remain for clinicians as well. Evidence suggests that transplant physicians consider QOL as secondary to the curative potential of HCT, underestimate patients' symptoms, and overestimate their QOL.^80,81  Increased awareness is needed regarding the impact of HCT on QOL and the importance of QOL to patients. Clinicians should use QOL literature to help their patients to make informed decisions.⁸²  In light of the literature reviewed here, we would consider the following as a framework for the counsel of patients considering HCT: First, in general, it is important to understand that allogeneic HCT is an arduous treatment that risks serious complications. Patients should be counseled on their disease-specific and treatment-specific risks for transplant-related morbidity and mortality, acute and chronic GVHD, infectious complications, primary disease relapse, and late complications of transplantation. They should be informed that these complications can result in worsened QOL. Second, the majority of studies reviewed here suggest the following positive outcomes: by 1 to 2 years after transplantation, greater than 60% of transplant survivors report their QOL as “good to excellent” and greater than 60% have no major functional limitations.^{16,22,26,29,37,48,52,83}  Third, available data suggest the following adverse outcomes: greater than 25% of survivors will have ongoing significant medical problems and greater than 25% experience emotional distress and impaired life satisfaction.^22,23,38  Stated differently, at least 25% of survivors have ongoing bothersome symptoms.^22,23,38  These summary statements, based on synthesis from this literature, can help facilitate an understanding of the clinical meaning and impact of this treatment and thereby help patients and their physicians come to informed decisions.

This work was supported at least in part by Cancer Center Support Grant 3 P30-CA7692-09 from the National Cancer Institute, National Institutes of Health, Bethesda, MD.

National Institutes of Health

Contribution: J.P. conducted literature search, analysis, and produced this paper; C.A. offered critical review of the paper; and H.J. contributed to the analysis and production of the paper and offered critical review.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Heather Jim, Moffitt Cancer Center, 12902 Magnolia Dr MRC-PSY, Tampa, FL 33612; e-mail: heather.jim@moffitt.org.