Modern treatment strategies have increased life expectancy in multiple myeloma, but little is known about the endocrine, metabolic and nutritional status of long-term survivors. We performed endocrine, metabolic, bone, body composition and nutritional evaluations in 32 patients with intensively-treated, advanced but stable, myeloma a median duration of 6 years from diagnosis and three lines of intensive treatment, including at least one haematopoietic SCT procedure. All patients were off active treatment. There was a high prevalence of endocrine dysfunction: hypothyroidism (9%), hypogonadism (65% males) and elevated prolactin (19%). Adrenocortical function was preserved despite large cumulative corticosteroid pretreatment. Biochemical markers were consistent with postmenopausal status in all females and infertility in males. Nutritionally, 59% were vitamin D insufficient/deficient, reduced serum folate in 25% and vitamin B12 in 6%. Total body DEXA scanning confirmed ‘sarcopenic-obesity’ in 65%, but reduced bone density was seen in a minority. We conclude that potentially correctable endocrine, metabolic and nutritional abnormalities are prevalent in heavily-treated patients with stable multiple myeloma. Preservation of bone supports the efficacy of bisphosphonate treatment from diagnosis, but sarcopenic-obesity may contribute to frailty. Ultimately, multi-system screening and appropriate interventions may optimise quality of long-term survival and further studies are warranted.
In previous decades, the average survival of patients with multiple myeloma was relatively short and treatment options limited. An evolving number of anti-myeloma treatments has resulted in significant increases in life expectancy over the last decade.1,2 Randomised controlled trials have established the role of combination chemotherapy followed by intensive chemotherapy and haematopoietic SCT (HSCT) as a standard of care in younger patients. The advent of ‘novel’ agents thalidomide, bortezomib and lenalidomide and emerging others, used alone or in combination, added further to the life expectancy. Significant numbers of patients are now surviving in excess of 5–10 years, consistent with a chronic disease state, the extension of survival being especially in the younger age groups. However, none of these evolving lines of treatment is free of significant toxicity, and although some treatment toxicities may be acute, reversible and short-lived, others may be insidious in onset and result in permanent or irreversible damage in various organ systems. Thus, as most patients are exposed to multiple agents in a lifetime with myeloma, toxicities are cumulative. Although such toxicities of individual drugs used as single agents may be well defined, in combinations their effects may be additive or synergistic with other cytotoxic and supportive agents, and to tissue/organ damage occurring by other means,3 including sites directly affected by the myeloma and by comorbidities commonly experienced by an ageing population. A late onset treatment toxic consequence (often referred to as a ‘late effect’) has been defined by the National Cancer Institute as ‘a health problem that occurs months or years after a disease is diagnosed or after treatment has ended. Late effects may be caused by cancer or cancer treatment. They may include physical, mental and social problems and second cancers’.4 Thus screening for these toxic treatment consequences is justified soon after treatment has been completed.
Therefore in the modern management of advanced myeloma, patients are compromised due to a complex set of pathological processes related to myeloma, its treatment and advanced age. The extended survival of increasing numbers of patients means that there is a need to consider the multiple organ dysfunctions that characterise the later phases of the disease. Although many of these issues are becoming increasingly recognised in patients with advanced myeloma, there are limited systematically collected data to define the health status, in particular endocrine, metabolic, bone and body composition abnormalities and the interaction with the underlying chronic disease state.
We propose that interaction of such a cluster of processes results in a unique syndrome, which may potentially benefit from dedicated management strategies and multidisciplinary care. To help define the pathophysiological features of patients with advanced myeloma which have arisen as a consequence of the combination of the disease itself, natural ageing and of the cumulative treatment consequences, we proceeded to characterise a spectrum of parameters including body composition, bone, nutritional and endocrine features arising in patients with advanced, heavily treated but stable myeloma.
Patients and Methods
Ethics and research governance
The study was conducted in compliance with the EU Good Clinical Practice directive and the ICH Good Clinical Practice, and formally approved by Sheffield Teaching Hospitals NHS Foundation Trust R&D Department following research ethics committee review. All subjects gave written informed consent.
Patients were recruited from outpatient clinics in the Sheffield Teaching Hospitals NHS Foundation Trust, UK after fulfilling eligibility criteria of symptomatic multiple myeloma, with initial treatment including induction chemotherapy consolidated by at least one HSCT procedure followed later by treatment with at least one additional line of treatment for progressive or relapsed disease. Importantly, in order to exclude the impact of active disease and acute toxicity of treatment, patients were in stable plateau phase with stable serological parameters within the preceding three months before study enrolment, and either off active cytotoxic re-induction treatment, or on maintenance treatment for at least three months. Thus, those with active disease were excluded.
This was a cross-sectional study. Information was collected regarding age and previous medical history including drug history. All subjects underwent assessments of height, weight, bone mineral density (BMD) and markers of hormone, nutritional and bone turnover status.
Blood was drawn in all but two subjects between 1000 and 1100 hours. Patients were non-fasting. Serum was dispatched to the local laboratory for routine assays on the day of collection. Remaining serum was frozen at −80 °C for later analyses.
Endocrine and fertility assays
Endocrine and fertility markers were measured in serum according to local GLP within our Hospital Trust on the day of collection. These included: thyroxine (free T4); thyroid stimulating hormone (TSH); cortisol pre- and 30 min post-synacthen stimulation; prolactin (Pr); total testosterone in men only; oestradiol (E2) and AMH (anti-Müllerian hormone) in women only; FSH (follicle stimulating hormone); LH (luteinising hormone); sex hormone binding globulin in men. Local reference intervals were used to establish out-of-range results. In men, free androgen index was estimated using the equation: 100 × (total testosterone/SHBG) to estimate the free (unbound) testosterone.5 A value of <30 indicates testosterone deficiency. In men inhibin B levels of <75 pg/mL was considered to be supportive of reduced spermatogenesis6,7 and in women AMH levels (which correlate with the follicular reserve) below the lower limit of laboratory measurement (<0.15 ng/mL) were interpreted as supportive of infertility.8
Serum vitamin D (25(OH)D), vitamin B12; ferritin and folate were measured according to local GLP and reference ranges on the day of collection.
Markers of bone turnover
Markers of bone turnover were measured in serum in a specialised bone metabolism laboratory. These included markers of bone formation (bone alkaline phosphatase and P1NP (type 1 procollagen amino-terminal propeptide) and βCTX (bone resorption C-terminal telopeptides)). Reference intervals for P1NP and βCTX were taken from Glover et al.9 These reference intervals were derived in young women during a period of optimal bone health and so are considered the best comparator. There are no reference intervals for men. Reference intervals for bone alkaline phosphatase are taken from Rogers.10
Body composition and BMD
BMD was measured at the total body, femoral neck and lumbar spine by DEXA (dual energy X-ray absorptiometry) using a Lunar Prodigy (GE Lunar Corp., Madison, WI, USA). Manufacturer’s precision of measurement for total body, lumbar spine and femoral neck is 1, 1.1 and 1.3%, respectively.
The total body DEXA was also used to collect body composition data including appendicular lean body mass and total fat mass in grams. Prevalence of sarcopenia (defined as ‘a loss of skeletal muscle mass and strength that occurs with advancing age’11) was estimated using previously published height- and weight-adjusted method and population-derived thresholds.12
The BMD was measured in grams per square centimetre and was weight-adjusted. BMD results were compared with an age-matched reference range provided by the manufacturer (Z-scores). The WHO definition for osteopaenia and osteoporosis was used where osteoporosis −2.5 s.d., osteopaenia (−1>−2.5 s.d.).13 In addition, we added the following categories: normal (>−1<1 s.d.), high (>1<2.5 s.d.), very high (2.5 s.d.).
Descriptive statistics were produced for all measured variables. For continuous variables this included calculation of standard summary statistics (median, range, standard deviation, minimum, maximum), whereas for categorical variables the percentages falling into each category were calculated. These were done for all the participants combined, and, where appropriate, by sex.
Demographics and treatment history
A total of 32 patients were recruited (17 males and 15 females), all having received at least one HSCT procedure (Table 1). A further 10 patients were approached, but they either declined to participate or were otherwise ineligible. The median age was 61 (range 41–71) years (women 62 years (range 51–73), men 61 years (range 41–74)). Median duration in years from diagnosis was 5.5 years (range 2–12) and 5.0 years (range 1.4–10) since their initial HSCT. Patients commenced treatment for symptomatic myeloma in the years 1998–2008. Patients had experienced a median of three (range 2–6) lines of treatment. All patients were Caucasian. Twenty-nine patients had at least one autologous HSCT with four having undergone allogeneic HSCT (three as first transplant and one as a sequential autologous-allogeneic procedure), and 10 had undergone two HSCT procedures. All patients had received cyclophosphamide in combination induction/re-induction and mobilisation regimens at doses of 1.5 g/m2 melphalan (up to doses of 140–200 mg/m2 in transplant conditioning) and high-dose steroids. Other anti-myeloma chemotherapy treatments are listed in Table 1. Two patients (6%) had received donor lymphocyte infusions after allogeneic HSCT. One patient also received TBI conditioning (8 Gy in one fraction). Thirteen patients had received at least one course of radiotherapy for palliative purposes with a median dose of 8 Gy in one fraction (Table 1). All patients had received recurrent courses of high-dose steroids, with an estimated median total dose of 2.3 g (range 0.4–5.1 g). At the time of assessment, twenty-nine (91%) patients were off all anti-myeloma treatment (except bisphosphonates) and three (9%) patients were receiving maintenance treatment with lenalidomide. All patients had received long-term bisphosphonate treatment, either daily oral sodium clodronate or monthly i.v. zoledronic acid, from diagnosis. At the time of the study, it was not the departmental practice to routinely provide vitamin D and calcium supplementation in all patients on bisphosphonate treatment, but to treat on an individual basis as required.
Eight (25%) patients gave a previous history of hypertension and one patient had a prior diagnosis of type II diabetes mellitus, but pre-existing cardiovascular diagnoses were infrequent (n=2).
Pre-existing endocrinopathies include one patient with hypothyroidism on thyroxine replacement, one woman on oestradiol hormone replacement therapy, three men with hypogonadism already receiving testosterone replacement and one patient was receiving fludrocortisone.
Endocrine testing identified an additional 2/32 (6%) patients with elevated TSH and one with low fT4. On ACTH (short Synacthen) testing, there was no evidence of adrenocortical insufficiency, with most patients having a clearly adequate response with serum cortisol level over 550 nmol/L; 3/32 had a borderline response but achieved an increment of >200 nmol/L above baseline following ACTH stimulation.
In addition to the three patients already receiving testosterone replacement for hypogonadism, eight men had low serum testosterone confirmed, that is, total 11/17 (65%) were hypogonadal. A total 10/17 (59%) men had a free androgen index less than 30. All males had inhibin B levels <75 pg/mL. In men with low testosterone, LH and/or FSH was raised in nine men and normal or low in eight men.
Date of last menstrual period was not recorded. Most women (14/15, 93%, age range 52–74 years, median age 63 years) had a biochemically confirmed menopause (that is, raised FSH and LH), of which one woman had oestradiol levels consistent with taking documented hormone replacement therapy. AMH levels in all these women were below the lower reference limit (<0.15 ng/mL) and consistent with low ovarian reserve. One patient, age 56, had a biochemical profile consistent with active ovarian function, with suppressed FSH and LH and a premenopausal oestradiol level (189 pmol/L). Despite this there had been no documented menses for over five years and AMH levels were consistent with infertility.
Elevated Pr levels were found in 6/32 (19%, two men and four women) with one patient having pathological level (1745 mIU/L), that is, greater than twice the upper normal range. This patient was reviewed by an endocrinologist but no underlying cause was identified and was attributed to fentanyl. Of the remainder with mildly elevated Pr, four patients were known to be on strong opioids including fentanyl and oxycodone and one on risperidone, which are known to elevate Pr. In one other patient, raised Pr was explained by compromised renal function.
Bone marker results indicated that 7/31 (23%) had raised bone ALP and 11/31 (36%) raised P1NP, both markers of bone formation. Bone resorption as measured by βCTX was elevated in 4/31 (13%) and reduced in 5/31 (16%). Table 2 gives the frequency of BMD by sites and distribution according to the WHO definition of osteoporosis, with additional categories added (see Patients and methods).
Body composition status
Body mass index was >30 kg/m2 (indicating clinical obesity) in 13/30 (43%, nine males and four females) and (25–29.9 kg/m2) in a further six patients (that is, clinically overweight). The arms of four very obese patients were not completely in the scan range (soft tissue only) meaning accurate recording of lean and fat mass were not available. These patients were excluded from sarcopenia analyses. Height and weight-adjusted sarcopenia was prevalent in 17/26 (65%, 7/12 males and 10/14 females) (see Patients and methods).
Nutritional testing indicated that 12/32 (37.5%) had deficient levels of vitamin D (that is, <30 nmol/L), and a further 7/32 (21.8%) had insufficient levels of vitamin D (30–50 nmol/L). Eight of 32 (25%) were deficient in folate. No patient had reduced ferritin, but levels were raised in 12/32 (38%) (none known to have a pre-existing diagnosis of hereditary haemochromatosis). Two of 32 (6%) had reduced levels of vitamin B12 (with no patient known to have pre-existing B12 deficiency).
In many cancers there is increasing evidence that patients are at risk of physical and psychosocial consequences of treatment. Observations of such ‘late-effects’ were originally applied to survivors of childhood cancer, and are increasingly relevant to adult survivors of cancer where treatment has been with curative intent.14, 15, 16, 17 However, there are other adult cancers, such as multiple myeloma, in which cure is rare, but where modern, frequently intensive, and repeated treatments may result in chronic disease control and prolonged survival. In these ‘chronic cancers’ it is becoming increasingly relevant to recognise the cumulative burden of the disease and treatment-related toxicity, and similar ‘late-effects’ screening protocols may thus be important in early detection and intervention of problems. Recent late-effects screening guidelines have been published for patients who have received allogeneic and autologous HSCT,18 but none exist specifically for patients with myeloma, who, in addition to HSCT, may accumulate side effects from other intensive and complex treatments, as well as ongoing damage from active myeloma.
We approached an otherwise unselected cohort of patients, who we considered ‘typical’ of younger patients in a UK tertiary haematological centre with myeloma, who had received HSCT as initial treatment and then treatment for one or more episode of progression or relapse, therefore defining a cohort of relatively heavily treated patients. Importantly, to maximally exclude the effect of active myeloma and acute treatment toxicity, we selected patients in stable plateau phase on the basis of paraprotein levels and being off active induction treatment for at least three months.
We have been able to detect abnormalities in endocrine, nutritional and metabolic parameters along with body composition in this relatively young and heavily treated cohort of myeloma patients. These abnormalities may have significant implications for ongoing physical and psychological health, although, importantly, may be amenable to simple screening and interventions.
In agreement with a previous report of myeloma patients treated with thalidomide before widespread use of novel anti-myeloma agents,19 around a tenth of patients in our cohort had evidence of thyroid hormone deficiency. In addition, pre-existing clinically confirmed hypothyroidism was observed in this population. Two patients had compensated hypothyroidism with mildly elevated TSH and normal fT4, which could be permanent but on a single measurement we cannot absolutely exclude a transitory state. In a further patient, low fT4 with normal TSH level appears to indicate secondary hypothyroidism, although the patient has not received cranial radiotherapy and as such this finding remains unexplained.
Despite the use of recurrent high-dose pulsed corticosteroid therapy in all patients, adrenal function was surprisingly well preserved. The significance of raised Pr levels (in around one fifth of our subjects) is best explained by drugs, especially strong opioids, in majority of patients.20
Male hypogonadism was observed in around two-thirds of men on the basis of a biochemical deficiency in testosterone, which, based on our previous studies,21 appears to be the most robust parameter for androgen deficiency in cancer survivors. In those patients with a low testosterone, nine had raised FSH suggesting primary hypogonadism and eight had normal or low gonadotrophins suggesting secondary hypogonadism. In line with our previous studies in other cancers,21,22 the frequent finding of hypogonadism may have significant physical and quality of life implications in long-term male survivors of myeloma and warrants further investigation.
In addition, this study provides new information regarding male fertility in survivors of myeloma. In this exploratory study involving older males, direct semen analysis was likely to have presented logistical and potentially ethical challenges, and therefore, we chose to use a surrogate marker. Van Beek et al.7 confirmed inhibin B to be an independent serum marker for spermatogenesis in younger men. Although our cohort is older, the universally low inhibin B levels are supportive of infertility in all males with intensively-treated advanced myeloma. As expected for the intensity of pretreatment and the age group, all men had surrogate biochemical evidence consistent with male infertility. Although ‘gold standard’ assessment should be pursued in individual cases as appropriate, the likelihood is that reproductive function is permanently lost in intensively-treated patients with myeloma.
The precise impact of intensive treatment of myeloma on female gonadal function was not as easy to define in a cross-sectional study, although all patients had ceased menstruation and virtually all patients were confirmed as biochemically postmenopausal. Prospective studies would be necessary to answer key questions in this area, although, as the average age of menopause is 51 years, this issue is only of potential relevance to a minority of female patients with myeloma.
This exploratory study in patients with stable myeloma provided an opportunity to examine patients who had been routinely treated with long-term bisphosphonates from diagnosis. Although there are limitations in bone densitometry in patients with fractures, based on total skeletal assessments and sites without fracture, the results supported normal-to-high levels of total bone density in most patients. Interestingly, despite clinical inactivity of the myeloma at the time of sampling, the biochemical bone markers were consistent with an ongoing state of increased bone turnover, although the non-fasting βCTX results need to be treated with caution.
In addition, we report for the first time a high prevalence of abnormal body composition, specifically ‘sarcopenia’, in patients with advanced myeloma. Despite a high prevalence of overweight and obese patients (that is, high BMI and fat mass) in this cohort, particularly in men, following adjustment for height and fat mass,12 there is a marked loss of muscle mass. The term ‘sarcopenic-obesity’, where fat mass remains increased but muscle mass is reduced12 may therefore be applied in the majority of patients in this cohort. Clinically, this observation is potentially important, indicating a state of muscle wasting despite the preservation or increase in fat mass (and lack of overall weight loss traditionally associated with advanced solid cancers). Presence of sarcopenia may add to the state of increased frailty,23 poor mobility and multifactorial disability commonly associated with patients with advanced stages of myeloma. Measurements of weight or BMI alone are thus poor proxies for sarcopenia and for assessment of body composition. Sarcopenic obesity is likely to be of insidious onset in long-term survivors of myeloma and the causes multifactorial, including endocrine, metabolic, nutritional, physical activity,23 along with myeloma itself and its treatment (cytotoxics and high-dose corticosteroids). Further systematic investigation is justified to define and characterise this problem and correlate its impact on muscle strength, frailty and other physical performance metrics, along with overall quality of life above and beyond that of normal ageing. Such information would be vital in evaluating effective interventions in this area in the long-term myeloma survivor.
A number of correctable nutritional deficiencies in the advanced myeloma patient have also been highlighted by this study, possibly explained by suboptimal diet, and insufficient exposure to sunlight, perhaps related to limited mobility. Many of these aspects are amenable to routine supplementation and/or screening. Folate deficiency was noted in 25% and reduced vitamin B12 levels were also relatively common (although spurious reduction in measured B12 is recognised in some patients with paraproteinaemia) which concurs with previous reports.24 Iron deficiency was not detected, but raised ferritin levels, best explained by previous transfusion, were common. In common with the general population25 and other myeloma studies,26 around one half of the patients had evidence of vitamin D insufficiency, potentially compounding myeloma bone disease. At the time of the study (before BCSH guidelines), it was not our routine practice to supplement patients with calcium and vitamin D supplementation, but there is clearly support for such supplementation, while monitoring serum calcium levels, not least to provide optimal levels for the prevention of fractures.27
This exploratory cross-sectional study is limited by its lack of control data and before and after treatment analyses. Longitudinal studies with serial sampling for assessment of endocrine, bone, metabolic and nutritional parameters from diagnosis through various treatments would help clarify these aspects, and potentially performed in conjunction with clinical trials of myeloma treatment with sufficient follow-up to permit assessment of ‘late effects’.
In summary, a number of novel observations and trends in a hitherto undercharacterised subpopulation of myeloma patients have been supported by the results of this study. These abnormalities are present in stable but advanced disease, and not explained by active myeloma or acute side effects of anti-myeloma treatment. This study is the first time that these aspects have been examined in the context of modern intensive management strategies which now enable increasing numbers of patients to survive multiple relapses of myeloma. Although firm conclusions from this study are limited by a lack of age-matched non-myeloma controls necessary to establish a clear excess in comorbidities, pragmatic recommendations for intensively-treated myeloma patients beyond first relapse might include routine screening/monitoring for endocrine abnormalities such as hypogonadism and hypothyroidism, along with vitamin D and folic acid insufficiency, which may add to the symptomatology and other problems associated with longstanding myeloma (for example, fatigue, bone disease) and are potentially correctable by simple replacement treatments. In addition, assessment of body composition changes may serve as a proxy for global health and frailty. Further research is necessary to define the problems, not only in intensively-treated younger patients, but also in less intensively-treated elderly patients. Ultimately, such evidence could add to updates of the current guidelines,1,2 be potentially invaluable in designing models for comprehensive survivorship care in myeloma, and provide a basis for interventional studies for improving outcomes.
The investigators thank Myeloma UK for supporting this research with the Richard Townley Research Grant. We also thank the patients for participating; Charlotte Morgan and the staff of the Clinical Research Facility at the Royal Hallamshire Hospital, Sheffield, UK for coordinating this research; Professor Richard Eastell, Consultant in Metabolic Bone Medicine for advice on bone marker measurement; Fatma Gossiel, Bone Metabolism Unit, University of Sheffield for measuring the bone markers; Sandra Gutcher and James Swinscoe for DEXA scans; the laboratory scientists and technicians at Sheffield Teaching Hospitals NHS Foundation Trust clinical laboratory. We also acknowledge the other members of the Late Effects Group, Sheffield.