SEARCH

SEARCH BY CITATION

Keywords:

  • birth cohort;
  • epidemiology;
  • extramedullary;
  • multiple myeloma;
  • survival

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

BACKGROUND.

The incidence of multiple myeloma (MM) is lower in Asia than in western countries. However, no data are available on descriptive epidemiology of MM in Chinese.

METHODS.

From 1979 to 2003, 3602 MM patients were registered in the Taiwan National Cancer Registry. The annual incidence and mortality were calculated and age-standardized to the world standard population in the year 2000. Age-period-cohort effects on incidence were analyzed. The salient clinical data of 526 MM patients in a single institute were also investigated.

RESULTS.

From 1979 to 2003, the average age-adjusted incidence per 100,000 population was 0.75. The incidence increased with age to a peak of 5.2 in those aged 75–79 years. In addition to age, remarkable period and birth cohort effects were found to contribute to increased incidence of MM. The age-adjusted mortality also increased, which accounted for an average of 0.59 per 100,000 deaths; however, the fatality rate was steady at 80%. Clinical and laboratory characteristics and treatment outcomes of the 526 MM patients were similar to those reported elsewhere. Remarkably, extramedullary myeloma (extra-MM) at diagnosis was more common in patients younger than 55 years of age than in others (43% vs 13%, P < .001).

CONCLUSIONS.

Incidence of MM in Taiwan has dramatically increased in recent years and is associated with a birth-cohort effect. There are no apparent differences in treatment outcome between MM patients in Taiwan and in other countries. However, prevalence of extra-MM is higher in patients younger than 55 years of age. Cancer 2007; 110:896–905. © 2007 American Cancer Society.

Multiple myeloma (MM) is a clonal plasma-cell neoplasm characterized by proliferation in the bone marrow (BM) of neoplastic plasma cells that impair hematopoiesis, activate osteoclastic bone resorption, and secrete a monoclonal paraprotein (M-protein) in serum and/or urine.1 MM accounts for about 1% of human neoplasms, almost 2% of deaths due to cancers, and 12%–15% of all cases of hematological malignancy.2, 3

The etiology of MM is unknown. Epidemiologic data suggest that age, genetic factors, chronic antigenic stimulation, and some environmental or occupational factors may play a role in pathogenesis of MM.1 Intriguingly, the variation in worldwide incidence is much greater for MM than for other hematologic malignancies.4 The reported age-adjusted incidence per 100,000 population of MM around the world is 0.5 in Hawaiian Japanese men, 0.9–3.3 in most European countries, and 8.2 in San Francisco Bay Area black men.5–7 The incidence of MM is much lower in the Chinese of Shanghai and Singapore than in the Caucasian populations of North America and Europe.2, 5 During the past several decades, the incidence of MM has increased in certain parts of the world,7–11 while it appears to have reached a plateau or even declined in some areas of the USA and UK since the 1990s.12–14 This dissimilarity has been the subject of considerable interest, and reasons for this rise in incidence of MM in these countries remain unclear.

To formulate a new etiologic hypothesis or to elucidate the mechanism for the inter-racial differences in incidence of MM, analyses of both the changes in MM geographic distribution and incidence over time may be helpful. Age-period-cohort (APC) analysis is commonly used by epidemiologists to analyze trends of disease incidence and mortality and to help develop hypotheses.15, 16 The age effect reflects physiologic differences between age groups in susceptibility to a given disease. The calendar time-period effect usually reflects factors that affect all age groups equally during a given period, such as introduction of new diagnostic or therapeutic techniques, and improvement in completeness of data registration. The birth cohort represents a population born into a specified generation. If disease incidence is affected by the birth cohort, then the term birth-cohort effect is used, which may involve factors that differentially affect birth cohorts because of differences in exposure levels, such as life style, risk, or protective factors.15–17 APC analysis is designed to estimate the effect of patients' age as well as individual effects of factors, such as time at diagnosis and patients' birth cohort, which are usually overlooked in cross-sectional studies.15, 16 At present, available data dealing with descriptive epidemiology of MM is mainly from western countries,1–5 and MM in Chinese has not been studied comprehensively. Moreover, APC analysis has rarely been used to analyze trends of MM incidence.

Our aims were to present descriptive epidemiology of MM in Taiwan, a country inhabited by 20 million Chinese located in the southeastern Asia, and to provide key epidemiological data in this population. The potential effects of patient age, calendar time period at diagnosis, and birth cohort on the change in incidence of MM in Taiwan for the past 25 years were examined. In addition, the salient clinical data of MM patients in one of the largest medical centers in Taiwan were investigated and reported herein.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Population

From 1979 to 2003, Taiwan had an average of 20 million inhabitants (men, 52%; women, 48%; almost exclusively Chinese), and the proportion of its aged population (older than 65 years) increased from 4% to 9%. In the year 2003, the life span of men and women was 73 years and 79 years, respectively.18

Cancer Registry

Hospitals island-wide are required to notify the Taiwan National Cancer Registry (TNCR), which was founded in 1979, of all new cases of cancer. This registry includes approximately 80% of all cancer cases in Taiwan.19 During the years 1979 to 2003, MM was coded according to the International Classification of Diseases for Oncology (ICD-O; World Health Organization, Geneva, Switzerland, 1976). The 1st Edition (ICD-O-1; 1979–1990), then the Field Trial Edition (ICD-O; 1991–2001), and the recent 3rd Edition (ICD-O; 2002–2003) coded MM as M-9730/3, M-9730/3, and M-9731/3-9734/3, respectively. MM mortality was coded as 203.00 by the ICD version 9 throughout the study period. Data for newly diagnosed cases and certified deaths due to MM were obtained from the Taiwan National Cancer Registry. Data for newly diagnosed cases of other hematological malignancies, including non-Hodgkin lymphoma (NHL) (ICD-O M-9590/3-9594/3, M-9670/3-9704/3), Hodgkin lymphoma (HL) (ICD-O M-9650/3-9667/3), acute myeloid leukemia (AML) (ICD-O M-9861/3, M-9862/3, M-9866/3), acute lymphoblastic leukemia (ALL) (ICD-O M-9821/3, M-9824/3), chronic myeloid leukemia (CML) (ICD-O M-9863/3) and chronic lymphocytic leukemia (CLL) (ICD-O M-9823/3) were also obtained for comparison.

University Hospital Patients

From 1966 to 2005, the data of 526 patients with MM diagnosed and/or treated at the National Taiwan University Hospital (NTUH) were analyzed. Criteria for diagnosis and staging of MM were defined according to Durie and Salmon.20, 21 All patients' laboratory reports were reviewed by our staff to verify diagnosis and staging of MM. Plasma cell leukemia (PCL) was defined as previously described.21 Patients with monoclonal gammopathy of undetermined significance, systemic amyloidosis, or other conditions associated with a monoclonal gammopathy, as defined previously,21 were excluded. Moreover, these hospital patients were grouped into 2 age strata: a group of ages ≤55 years and a group of ages >55 years. The cutoff of 55 years was chosen because this is the age limit for high-dose chemotherapy followed by autologous hematopoietic stem cell transplantation (HDC/ASCT) at this center.

Treatment and Treatment Response

Among the 526 university hospital patients, 446 (85%) were treated with MP (melphalan and prednisolone), or CP (cyclophosphamide and prednisolone), or VAD (vincristine, doxorubicin, and dexamethasone), as described previously22; the other 80 (15%) were not chemotherapy recipients. Among recipients of chemotherapy, 30 also received HDC/ASCT.

Criteria for treatment response were defined by EBMT (European Group for Blood and Marrow Transplant) guidelines.23 Progression-free survival (PFS) was defined as the interval between initial obtainment of at least minimal response (MR) or better and documented relapse or progression. Overall survival (OS) was defined as the period between the date of diagnosis and the date of death for any reason.

Statistics

Age-specific, sex-specific, and calendar year-specific incidence and mortality were estimated by using the number of newly diagnosed and mortality cases, respectively, in each age, sex, and calendar-year group as the numerator, with corresponding annual resident midyear population as denominators.18 Fatality rate (percentage) was 100 times the mortality divided by the corresponding incidence. The values of incidence and mortality were expressed as rates per 100,000 persons, and these rates were adjusted for age by the direct method on the basis of the world standard population in the year 2000.17

These annual data were averaged into 5 5-year time periods (from 1979–1983 to 1999–2003). Data from patients younger than 30 years and older than 85 years were omitted because of the low case number. For the APC analysis, data were categorized into 11 5-year age groups (30–34 years to 80–84 years) and 5 5-year time periods (from 1979–1983 to 1999–2003), suggesting 15 overlapping 9-year birth-cohort groups (from midyear 1899 to 1969).

The chi-square or Fisher exact tests were used for between-group comparison of discrete variables. The 2-sample Student t test was used for between-group comparison of means. Kaplan-Meier survival curves were used for estimation of PFS and OS. On survival analysis, those patients who had received HDC/ASCT were analyzed separately. All directional P values were 2-tailed, with a P ≤ 0.05 considered significant for all tests. All analyses were performed by SPSS 11.0 software (SPSS Institute, Chicago, Ill).

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

Incidence

From 1979 to 2003, a total of 3602 MM patients, including 2328 (65%) men and 1274 (35%) women were registered in the Taiwan National Cancer Registry, and the average incidence was 0.75. The average age-specific and sex-specific incidence of MM for 2 10-year periods, 1979–1988 and 1994–2003, is summarized in Table 1.

Table 1. Sex-Specific and Age-Specific Incidence of MM in Taiwan (per 100,000 Population)
Age, y1979–19881994–2003
Men (95% CI)Women (95% CI)Both sexes (95% CI)Men (95% CI)Women (95% CI)Both sexes (95% CI)
  1. CI indicates confidence interval.

30–390.05 (0.01–0.09)0.04 (0.01–0.06)0.05 (0.03–0.06)0.17 (0.12–0.22)0.08 (0.03–0.13)0.13 (0.10–0.15)
40–490.34 (0.23–0.46)0.24 (0.11–0.38)0.30 (0.22–0.37)0.74 (0.61–0.87)0.37 (0.27–0.48)0.56 (0.47–0.65)
50–591.46 (1.17–1.75)0.97 (0.71–1.23)1.24 (1.02–1.47)2.82 (2.45–3.19)1.45 (1.21–1.70)2.14 (1.85–2.42)
60–692.77 (2.32–3.23)1.77 (1.31–2.23)2.34 (1.94–2.74)5.87 (5.43–6.30)4.60 (4.13–5.08)5.26 (4.95–5.57)
70–792.97 (1.88–4.05)1.16 (0.70–1.61)2.07 (1.39–2.75)10.47 (8.93–12.02)6.28 (5.13–7.44)8.59 (7.35–9.82)
80+2.01 (0.32–3.70)0.17 (−0.16–0.49)0.91 (0.40–1.43)9.70 (7.34–12.05)4.47 (2.37–6.58)6.95 (4.73–9.18)

For both sexes, the average incidence in 1979–1983 and 1999–2003 was 0.36 and 1.21, respectively (Fig. 1A), whereas that in men was 0.46 and 1.48, respectively (Fig. 1B), and that in women was 0.25 and 0.93, respectively (Fig. 1C). For comparison, the average incidence of other hematologic malignancies during the study period is also shown in Figure 1. For both sexes, the fold increase in incidence of MM and other hematologic malignancies (between 1979–1983 and 1999–2003) was greatest for MM (3.4 fold), followed by CLL (3.3), NHL (3.2), CML (1.7), AML (1.6), ALL (1.5), and HL (1.4) (Fig. 1A). In 2003, MM accounted for approximately 1% of all cancers in Taiwan, and the incidence of MM exceeded that of ALL, especially in men. Therefore MM is now the third most common hematologic malignancy in Taiwan, just behind NHL and AML (Fig. 1). From 1979 to 2003, the sex ratio (M:W) of incidence of MM decreased gradually from 3.6:1 to 1.5:1, with an average of 1.8:1. The decrease in sex ratio was quite steady in all age groups.

thumbnail image

Figure 1. Average age-adjusted incidence of MM and other hematological malignancies in Taiwan (1979–2003), by sex and 5 5-year calendar time periods. (A) Both sexes; (B) men; (C) women. Rates are plotted by using a logarithmic scale. NHL indicates non-Hodgkin lymphoma; AML, acute myeloid leukemia; ALL, acute lymphoblastic leukemia; MM, multiple myeloma; CML, chronic myeloid leukemia; HL, Hodgkin lymphoma; CLL, chronic lymphocytic leukemia.

Download figure to PowerPoint

Figure 2 shows the age-period-cohort effects on MM incidence. In Figure 2A, incidence of MM increased for all age groups but was greater in patients ≥60 years than in those <60. Notably, the age of peak incidence shifted to the right in later time periods. In Figure 2B, at any given age, incidence increased linearly over time and more steeply in the older age groups (≥70 years). Notably, between 1984–1988 and 1989–1993, the incidence increased more consistently at any given age than in other time periods. Figure 2C shows that the incidence was higher for later birth cohorts in every given age group, and, particularly for those older than 60 years, incidence rose dramatically in successive birth cohorts.

thumbnail image

Figure 2. Incidence of MM in Taiwan, 1979–2003. Plots show (A) differences in age-specific rates between 5 selected calendar periods, (B) secular (calendar) trends in age-specific rates, (C) cohort trends in age-specific rates. Rates are plotted by using a logarithmic scale.

Download figure to PowerPoint

Mortality and Fatality

The average sex-specific and age-specific mortality for 2 different calendar periods (1979–1988 and 1994–2003) are summarized in Table 2. The average mortality (1979–2003) was 0.59 (0.73 for men and 0.42 for women), and the average fatality rate remained fairly constant (81% for men and 80% for women). However, for both sexes, the fatality rates were lowest at ages 35–44 years, relatively steady at ages of 45–59 years, gradually increased by age 60 years, and peaked at ages older than 80 years. Interestingly, the fatality rates were also high at ages 30–34 years, especially in women (Fig. 3A). Age-specific fatality rates between 1979–1988 and 1989–2003 decreased in nearly all age groups, except in the 55–70 years of age group (Fig. 3B).

thumbnail image

Figure 3. The fatality rate of MM patients in Taiwan. The fatality rates, presented as percentages, are shown for (A) men (solid line) and women (dotted line), and for (B) 2 calendar time periods (1979–1988, dotted line; 1994–2003, solid line).

Download figure to PowerPoint

Table 2. Sex-Specific and Age-Specific Mortality Due to MM in Taiwan (per 100,000 Population)
Age, y1979–19881994–2003
Men (95% CI)Women (95% CI)Both sexes (95% CI)Men (95% CI)Women (95% CI)Both sexes (95% CI)
  1. CI indicates confidence interval.

30–390.03 (0.00–0.05)0.04 (0.00–0.08)0.03 (0.02–0.05)0.07 (0.04–0.09)0.04 (0.01–0.07)0.06 (0.03–0.08)
40–490.20 (0.14–0.26)0.13 (0.07–0.19)0.17 (0.12–0.21)0.43 (0.26–0.60)0.21 (0.12–0.31)0.32 (0.22–0.42)
50–591.30 (1.01–1.58)0.62 (0.35–0.89)1.01 (0.83–1.19)1.73 (1.45–2.01)0.95 (0.71–1.20)1.34 (1.10–1.58)
60–691.84 (1.45–2.23)1.43 (1.01–1.84)1.66 (1.32–2.00)4.78 (4.18–5.38)3.25 (2.79–3.71)4.03 (3.54–4.51)
70–792.62 (1.49–3.76)1.23 (0.67–1.79)1.94 (1.23–2.56)8.37 (6.87–9.87)5.78 (4.77–6.79)7.20 (6.00–8.39)
80+3.28 (0.86–5.69)0.64 (−0.11–1.39)1.69 (0.56–2.82)12.25 (10.06–14.44)5.16 (3.65–6.67)8.55 (7.01–10.08)

Clinical Characteristics

Clinical and laboratory features of 526 patients are summarized in Tables 3 and 4. Up to June 2006, with a median follow-up of 100 months, 421 (80%) patients were lost to follow-up because of either death (347 patients) or other reasons (74 patients). Among the 347 MM patients who died, the most common causes of death were pneumonia or sepsis (57%) and myeloma progression (25%).

Table 3. Salient Clinical Characteristics of the 526 MM Patients Treated at NTUH
CharacteristicsAll Patients (N = 526)Age ≤55 y (n = 164)Age >55 y (n = 362)P
  • *

    Mean ± SD.

  • Manifestation of extramedullary myeloma at diagnosis (see text).

Age at diagnosis, y*62.0 ± 12.846.9 ± 7.368.8 ± 8.0 
Sex, M/W344/182112/52232/130.374
Major presentation (%)
 Bone pain288 (55)88 (54)200 (55).777
 Dyspnea065 (12)19 (12)046 (13).776
 Tumor formation038 (7)22 (13)016 (4)<.001
 Fever & infection033 (6)04 (2)029 (8).018
 Health check-up028 (5)09 (6)019 (5)1.000
 Others, unspecified074 (14)22 (13)052 (14).892
Complication (%)
 Pathological fracture164 (31)42 (26)122 (34).068
 Acute renal failure052 (10)19 (12)033 (9).431
 Neuropathy029 (6)21 (13)008 (2)<.001
 Amyloidosis012 (2)04 (2)008 (2)1.000
Durie-Salmon stage (%)   .703
 I & II218 (41)70 (43)148 (41) 
 IIIA & IIIB308 (59)94 (57)214 (59) 
Extra-MM (%)117 (22)71 (43)046 (13)<.001
Table 4. Laboratory Results and Treatment Response of MM Patients Treated at NTUH
CharacteristicsAll patients (N = 526)Age ≤55 y (n = 164)Age >55 y (n = 362)P
  • BM indicates bone marrow; CRP, C-reactive protein; HB, hemoglobin; HDC/ASCT, high dose chemotherapy followed by autologous hematopoietic stem cell transplantation; VAD, vincristine, doxorubicin, dexamethasone; β2; M, β2-microglobulin.

  • *

    Mean ± SD.

  • Including IgM, unspecified M-proteins, and nonsecretory MM.

  • Number of patients who had received chemotherapy; responders included patients who had minimal response, partial response, or complete response; nonresponders denoted patients who had no change or progressive disease (see text).

Immunoisotype (%)   .759
 IgG253 (48)74 (45)179 (49) 
 IgA127 (24)36 (22)91 (25) 
 Light-chain103 (20)34 (21)69 (19) 
 IgD16 (3)7 (4)9 (3) 
 others27 (5)13 (8)14 (4) 
κ:λ ratio1.3:11.1:11.4:1.277
HB, gm/dL*9.2 ± 2.69.7 ± 2.99.0 ± 2.4.006
Albumin, gm/dL*3.3 ± 0.73.4 ± 0.73.2 ± 0.7.005
Creatinine, mg/dL*2.3 ± 2.32.1 ± 2.22.4 ± 2.3.250
CRP, mg/dL*2.7 ± 4.42.0 ± 2.63.0 ± 4.9.048
β2M, mg/L*7.9 ± 10.26.6 ± 8.08.5 ± 10.9.137
Plasma cell, BM, %*43.0 ± 28.242.7 ± 30.543.1 ± 27.1.876
Chemotherapy446146300 
 VAD220100120 
Evaluable patients392131261 
 Responders (%)210 (54)72 (55)138 (53).899
 Nonresponders (%)182 (46)59 (45)123 (47) 
HDC/ASCT30291 

Neuropathy was noted in 29 (6%) patients at diagnosis. In this group, compartmental neuropathy for emerging plasmacytoma occurred in 19 (65%) patients, and idiopathic polyneuropathy was observed in 10 (35%) patients. MM was diagnosed in 12 (2 %) patients with amyloidosis, lytic bone lesions, and >30% plasma cells in the bone marrow. It was noteworthy that 117 (22%) patients overall had extramedullary manifestations of myeloma (extra-MM), including plasmacytomas in 92 patients, malignant effusions in 25 patients, and PCL in 12 patients. Importantly, on light-chain analysis, kappa-type predominance occurred for all immunoisotypes as would be expected, except for IgD MM, where 14 of 16 (88%) were lambda in type. Ten patients had bi-clonal and 1 patient had tri-clonal M-protein production.

Age was ≤55 years for 164 patients and >55 years for the remaining 362 patients. Clinical features and laboratory results for the 2 age groups are compared in Tables 3 and 4. The prevalence of extra-MM at diagnosis was significantly higher in patients ages ≤55 years than ages >55 years. The mean age was also significantly lower for those with than without extra-MM at diagnosis (55.5 ± 14.0 years vs 63.8 ± 11.8 years, P < .001).

Treatment Response and Outcome

Of the 446 patients treated with conventional chemotherapy, 392 were assessable for response (Table 4). There were 210 (54%) patients who responded, including 13 (6%) who had CR, 124 (59%) who had PR, and 73 (35%) who had MR. The median PFS was 19.0 ± 2.1 months. Patients ages ≤55 years and ages >55 years did not differ significantly in treatment response rate and PFS (median, 20.0 ± 6.0 months vs 19.0 ± 2.4 months, P = .622). For patients who responded to conventional chemotherapy, the median for the relative overall survival and 5-year and 10-year survival were 44.0 ± 3.3 months, 36%, and 10%, respectively; in contrast, for those who did not respond, the median for the relative overall survival and 5-year and 10-year survival were median 19.0 ± 2.9 months, 15%, and 3%, respectively.

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

To our best knowledge, this is the first report to comprehensively describe the epidemiology of MM in Chinese. We report here an almost 4-fold increase in the incidence of MM in Taiwan during the past 25 years, although the overall incidence of MM here is still quite low compared with that in western countries.1–7 Of the incidences of all hematological malignancies, that for MM has increased the most. These findings are comparable to those of other recent reports.8, 24

Reasons for great racial disparity in incidence of MM are unclear. However, 1 or more genetic factor(s) is likely to be playing some role. The incidence of MM is similar in Asian immigrants to the USA and Australia and in Asians in their countries of birth, but it is lower than the incidence in those born in the USA and Australia.2, 25 To support the notion of genetic influence, familial risk of MM has also been demonstrated.26, 27 In addition, a B-allele polymorphism of the tumor-suppressor gene encoding poly(adenoside diphosphate [ADP]-ribose) polymerase has been reported in patients with MM and monoclonal gammopathy of unknown significance (MGUS).28 Its expression was more frequent in US blacks, who have the highest incidence of MM worldwide, than in whites.1, 29 Notably, frequency of this B-allele polymorphism is much lower in Chinese than in whites.29 Moreover, frequency of several MM-related human leukocyte antigens (HLA), such as HLA-B5, HLA-Cw2, and HLA-C5,27 is also much lower in the Taiwan general population and in US Asian populations than in other US populations.30, 31 Thus, racial differences in genetic background may partly explain the lower incidence of MM in Asian countries. However, further studies, particularly in the area of molecular epidemiology, are warranted to uncover the genetic determinants of MM.

Our data suggested that, in addition to an age effect, both period and cohort effects contributed to the increasing trend in incidence of MM. The incidence increased more consistently between 1984–88 and 1989–93 at any given age than other time periods, which suggested a dominant period effect during this time period. Multiple factors may be involved in the period effect, such as improvement in diagnostic techniques and case ascertainment.12 Interestingly, the use of immunological methods, such as protein electrophoresis and immunofixation electrophoresis, became popular in Taiwan during this time period, supporting this view. Not unlike the reports of Kyle et al13 and Pollan et al,32 a birth-cohort effect on the incidence of MM was observed in this study. The stronger effect on incidence of MM in the older birth cohort (before 1934 birth cohort) indicated that changes in occupational or environmental exposures between patients born in earlier and later cohorts likely plays some role in the pathogenesis of MM. One of the possibilities is that the occupations of the older birth cohort, which were predominantly related to agriculture, were gradually switched to those of the younger birth cohorts, which were related to industry. Employment in agriculture has been most frequently associated with MM.33 Conversely, the prolonged birth-cohort effect observed in this study suggests that lifestyle or persistence of some environmental pollutants (such as polychlorinated biphenyls [PCBs], polychlorinated dibenzo-p-dioxins, and polychlorinated dibenzofurans [PCDFs]) may play a role in the pathogenesis of MM.34 Intriguingly, results from a global survey of such pollutants in the muscle of skipjack tuna showed higher levels in the East and South China sea, where Taiwan is located.35 Although the relations among toxins, food sources, environmental pollution, and increasing incidence of MM are speculative, these findings provide rationale for further analytical epidemiology studies to test this hypothesis.

One limitation of the analysis is the difficulty in differentiating a time-sustained period effect from a cohort effect. Because it is probable that access to modern hospital care improved evenly in Taiwan, the birth-cohort effect reported here could have been partly influenced by improved detection. However, the effect of exposures known to be associated with MM incidence, including radiation, agricultural pesticides, dioxin, and other related herbicides, may have been more intense for recent birth cohorts. Furthermore, the data shown in Figure 2A may be, in some degree, suggestive of a time trend in diagnostic bias, as a result of an historical under-recognition of MM in the oldest ages, which had steadily lessened and was largely ameliorated in the most recent calendar period. Therefore, the consequences of this bias would be associated with inflation of the overall estimate of increase in MM rates and the appearance of a greater increase in rates for the older ages. Undoubtedly, the age-differential diagnostic bias was an alternative explanation for the time trend of increased incidence of MM. However, it is important to note that the increased rate in older patients (>70 years; Fig. 2A) was more than 10-fold during this 25-year period, and this, therefore, seems difficult to explain by diagnostic bias alone. In addition, the observed increased incidence of MM in successive generations at any given age (not limited to older ages; Fig. 2C) also argues against the improved case ascertainment to be the only determinant for increased incidence of MM.

In contrast to other reports,10 sex ratio of MM in Taiwan decreased gradually from 1979 to 2003; and it was quite steady in all age groups. Thus, although MM incidence continued to increase in both genders, MM incidence increased more in women than in men during this period. The reason is not clear. In countries with reportedly higher rates of MM, such as Nordic countries, M:W ratios tend to be higher in younger than in older age groups, perhaps suggesting that exposure to occupational factors may affect men at an earlier age than it does women and lead to an earlier appearance of MM.36 In contrast, the fixed sex ratio in nearly all age groups in our data reinforces the hypothesis that exposure to environmental pollutants rather than occupational factors likely triggers increased incidence. A survey performed in Taiwan found higher levels of pollutants (eg, PCBs and PCDFs) in older than in younger persons and in women than in men.37 Furthermore, during 1956–1975, the average proportion of agriculture-related occupations was not much different between men and women (40% vs 46%, respectively) registered in our National Labor Data Base (Available at: http://www.dgbas.gov.tw).

Our study found that, in young age groups, fatality rates were higher in women than in men, and the reason is not clear. The decreased fatality rates in nearly all age groups, except those between 55 and 70 years, may reflect improved supportive care and tumor treatment, such as HDC/ASCT, which has been shown repeatedly to provide better outcome than conventional chemotherapy in MM patients, especially in younger MM patients.38, 39 In concordance with this observation, by a risk-adjusted retrospective analysis, our treatment outcome was better for our MM patients who had undergone HDC/ASCT than those who had conventional chemotherapy alone.

Although the incidence of MM is low in Taiwan, the clinical presentations and treatment outcomes of our MM patients were comparable to those reported in other countries,40, 41 which indicates a common pathogenesis of MM worldwide. Most of our patients with neuropathy reported numbness on extremities, and, in part, this was a result of compartmental neuropathy from emerging plasmacytomas. This finding may explain why patients younger than 55 years had a higher incidence of neuropathy in this study because they had higher incidence of extramedullary plasmacytomas than others. Nonetheless, the reported rate of symptomatic peripheral neuropathy was significant, and it appeared to be in the range of approximately 10%, which is somewhat higher that has previously been seen. The higher incidence of infections among patients older than 55 years was probably due to age and some concomitant disorders. Intriguingly, the incidence of extra-MM at diagnosis was significantly higher in patients ages ≤55 years than in patients ages >55 years. A similar observation has been reported in several cases of MM in younger patients.42, 43 Interestingly, frequent extramedullary relapse after HDC/ASCT has been reported in Spanish MM patients of median age 53 years, but the underlying mechanism(s) remained unclear.44 Therefore, studies are needed to elucidate whether the pathogenesis or clonal evolution of MM in younger patients is different from that in older patients. This is of importance, as, clinically, some emerging treatment strategies and novel drugs, such as bortezomib, have been shown to be especially active in such patients with extra-MM.45

In conclusion, the incidence of MM is indeed lower in Taiwan than in western countries, which may be because of genetic disparity between races. The incidence of MM continues to increase, and not only age and calendar period but also birth cohort is contributing to this increase. The clinical features and treatment outcome of our MM patients are similar to those of others. Intriguingly, extra-MM at diagnosis is more prevalent in MM patients younger than 55 years, and reasons for this finding need to be further evaluated.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES

We thank Ms. Mo-Lan Huang, Yi-Xuang Yuang, Li-Hwei Lin, and the staff of the national and hospital cancer registries for their help in acquiring data used in this study.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. REFERENCES