Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Incidence and management of hepatocellular carcinoma among Māori and non‐Māori New Zealanders

Incidence and management of hepatocellular carcinoma among Māori and non‐Māori New Zealanders H epatocellular carcinoma (HCC or primary liver cancer) is one of the most common cancers worldwide. Annual global estimates range from 500,000 to 800,000 new cases. The Asia‐Pacific region is believed to contribute up to half of these, a number that reflects a high prevalence of major risk factors. Chronic hepatitis B and chronic hepatitis C account for 80% of primary liver cancer cases worldwide. Other risk factors include heavy alcohol consumption (particularly alcoholic cirrhosis) and dietary aflatoxin. Infant hepatitis B vaccination is decreasing the rate of chronic hepatitis B‐attributable HCC internationally. However, the number of cases of HCC globally is expected to treble by 2050 because of increasing chronic hepatitis C infection and non‐alcoholic fatty liver disease. In NZ, liver cancer is an important health issue, with 253 new registrations in 2009, and a high rate of mortality with 208 deaths recorded in the same period. Large ethnic disparities in incidence have previously been described, with high rates in Māori, Pacific and Asian populations. Blakely et al. found that Māori had seven times the rate of HCC compared with non‐Māori/Pacific/Chinese populations (in 1983–1994). There is insufficient evidence to say whether these disparities are changing over time. Mortality from HCC is also considerably higher for Māori. Furthermore, there is some evidence that survival for Māori with HCC may be lower than for non‐Māori. Survival from HCC is determined by the stage of cancer at diagnosis, with earlier stage cancers associated with improved outcomes. This is partly due to tumour biology, but is also related to whether aggressive therapy can be undertaken with curative intent (currently surgical resection and hepatic transplantation). Likelihood of survival is also determined by the presence and severity of associated hepatic cirrhosis, which is a risk factor for tumour development and reduces hepatic reserve, compromising the range of anti‐cancer therapies that can be employed. There is evidence in NZ from other cancer contexts that treatment differences may impact on disparities in outcomes, but this has not been studied for HCC. This study examined the incidence of HCC in Māori and non‐Māori New Zealanders over a 24‐year period (1981–2004) in more depth than previously undertaken, to investigate time trends in disparities. We reviewed the clinical notes of a sample of Māori and non‐Māori patients with HCC to determine differences in treatment pathways, comorbidity and care, which may affect disparities in outcomes. Methods There are four main ethnic groups in NZ. According to the 2006 census, 65% of the population identify as European, 14% as Māori, 7% as Pacific, and 9% as Asian, with around 11% identifying with other ethnic groups. This paper focuses on comparing those from the indigenous Māori population with those of other ethnicities. This paper draws upon two datasets. Incidence trends have been calculated using data from a national data linkage study (‘Cancer Trends’) and data on treatment pathways is drawn from clinical notes review. The methods of each component will be outlined separately. Cancer trends The New Zealand Cancer Registry (NZCR) is a population‐based register that collects information on all primary malignant tumours diagnosed in NZ, excluding squamous and basal cell skin cancers. Ethnicity data held by the NZCR is of variable quality, introducing numerator‐denominator bias in ethnic‐specific incidence rates that varies over time. For this reason, in studies investigating ethnic trends in cancer incidence, using census records linked to the NZCR is preferred. Five closed cohorts were created of the NZ usual resident population (25 years and older) on census night 1981, 1986, 1991, 1996, 2001, followed up for incident cancer(s) until the subsequent census or, in the case of 2001 cohort, until 31/12/2004. Cohorts were created using probabilistic record to anonymously link census and cancer register records within a geographic area (meshblock or census area unit) on sex, date of birth, ethnicity and country of birth. A modified total ethnicity approach was used. This includes individuals in all the ethnic groups that they identified with and leaves a residual non‐Māori/Pacific/Asian group, titled European/Other. For the purposes of this paper, Pacific and Asian populations have been excluded because of their small numbers (an average of about 15 Pacific and 10 Asian liver cancer cases per year over the study period). The linkage methods, processes and outputs for Cancer Trends have been detailed elsewhere. Briefly, 73.2% (1981–1986) to 81.7% (2001–2004) of eligible cancer registry records were linked to a census record, with 95.2% (1981–1986) to 96.9% (2001–2004) of these linked census‐cancer records estimated to be true links. Liver cancer, as it pertains to the Cancer Trends study, is defined as primary liver cancer (ICD code C22) and consists principally of hepatocellular carcinoma (the more common type), as well as cancer of the intrahepatic bile ducts (peripheral cholangiocarcinoma). Incidence rates and rate ratios (with 95% confidence intervals) were calculated after direct standardisation of the cohorts to the age structure of the WHO World Standard population. Statistical tests of trend were conducted for rates and of the log transformed rate ratios. All measures were also calculated for all five cohorts pooled. All analyses were conducted in SAS v9. Notes review Incident cases of HCC (ICD code C22.0) diagnosed between 01/01/2006 and 31/12/2008 were identified from the NZCR. Patients eligible for study inclusion were those aged 25 years or older at diagnosis, normally resident in NZ, who had not received a prior diagnosis of HCC and were diagnosed prior to death. Because more than 90% of Māori patients lived in the North Island, the notes review was limited to this area. All eligible Māori patients living in NZ's North Island along with a randomly sampled equal number of non‐Māori patients were included. Ethnicity information was drawn from the NZCR. Patients were classified as Māori if they had been identified as Māori on any previous health record. Clinical data were obtained from patients' medical records in both public and private hospitals. Data were recorded on a standardised study pro‐forma by a trained oncology nurse. Relevant data were double‐entered and discrepancies resolved. Data included: patients' symptoms on presentation; surveillance status (whether they were being surveyed for development of HCC); viral serology (HBV or HCV positive); tumour characteristics (tumour grade, size and stage at diagnosis [AJCC TNM staging system]); receipt of treatment (surgical, radiotherapy and chemotherapy); details of surgical intervention (type of surgeon, hospital and procedure as well as any pursuant complications); adjuvant care provided (reoperations, other operations); timing of treatment; the cancer treatment pathway (including referral, review, offering and receipt of treatments); and palliative care (referral and receipt). Mortality data were derived from the national Mortality Collection, with patients followed until 31/12/2010. Survey methods were used to create population estimates for the total NZ cohort of liver cancer patients over the study period. The final Māori and non‐Māori samples were weighted to the total eligible NZ Māori and non‐Māori liver cancer populations. Age‐standardised rates were calculated by direct standardisation using the total NZ cancer population. This adjusts for the differences in age structure between Māori and non‐Māori. Differences in trends between Māori and non‐Māori populations were calculated using Cochran‐Mantel‐Haenszel chi‐squared tests, stratified by age group. Survival trends were investigated using Cox proportional hazard regression models, adjusted for age (continuous variable), sex (male/female) and stage (I‐IV). Follow‐up time started at date of diagnosis and ended at date of death from liver cancer. Patients who died from other causes were censored at that date, and those who did not die were censored at 31/12/2010. All analyses were performed using SAS Version 9.2. This study was granted approval from the Multi‐Regional Ethics Committee (MEC 10/042/EXP). Results Cancer trends Figure shows the markedly higher incidence rates of HCC among Māori over all time periods and for both sexes. In males, the pooled (over time) annual rate for Māori of 19.6 per 100,000 (17.1–22.1) was nearly five times the rate for European/Other (RR=4.79; 95% CI 4.14–5.54). The pooled rate for Māori females was 6.1 (4.6–7.6) per 100,000 women and for European/Other females was 2.1 (1.9–2.3) per 100,000 women (RR=3.02; 95% CI 2.33–3.92). Standardised rates of liver cancer for 25+ year old patients per 100,000 per year. For both ethnic groups, the incidence of liver cancer increased between 1981 and 2004: by 90% for Māori males ( p for trend 0.03) and 79% for European/Other males ( p for trend <0.01), see Figure a. For European/Other females, incidence increased 29% ( p for trend 0.07) but Māori females' incidence rates were unstable over time (Figure b). The corresponding standardised rate ratios for Māori compared to European/Other ranged from 4.38 (2.79–6.88) in 1981–1986 to 5.08 (3.58–7.21) in 1986–1991 for males, with no evidence of increasing inequalities on a relative scale over the time period studied ( p for trend 0.90). Similarly, there was no evidence of increasing inequalities among females, with standardised ratios for Māori compared to European/Other varying between 4.69 (2.47–8.89) in 1981–1986 to 2.10 (1.03–4.31) in 1986–1991. However, in absolute terms the standardised rate difference did show a trend towards widening inequalities for males, increasing from 10 (5 to 15) to 19 (13 to 24) per 100,000 from 1981–1986 to 2001–2004 ( p for trend 0.04). No such trend in absolute inequalities was evident for females. Notes review There were 440 registrations for the ICD code C22.0 (HCC) in NZ during the study period (diagnosed 1/1/2006 to 31/12/2008). In the North Island, there were 117 Māori patients, 317 non‐Māori patients and six patients with missing ethnicity data (these were later merged with the non‐Māori cohort prior to random selection). After exclusion criteria were applied, there were 103 Māori diagnosed with liver cancer over the period studied in the North Island. This cohort was included along with a random sample of 103 from the 200 study‐eligible non‐Māori, also from the North Island. Following clinical notes review, 9% of patients were found ineligible for study inclusion. This comprised eight patients found to have been diagnosed outside of the study period, four patients with an element of cholangiocarcinoma (mixed tumours) to their diagnosis, three diagnosed and treated in Australia, one found to have metastatic liver disease, one with no evidence of a cancer diagnosis and one non‐resident. This resulted in a final sample of 97 Māori and 92 non‐Māori patients. The ethnicity division of the non‐Māori sample included 75% European ethnicity, 13% Pacific ethnicity and 11% Asian ethnicity. Table shows the weighted, crude and age‐adjusted characteristics of the 189 patients that comprise the final cohort. Around 80% of patients were men in both ethnic groups. Māori were diagnosed at a younger age than non‐Māori (mean age at diagnosis for Māori 58.6 years compared with 65.4 years for non‐Māori; p <0.001), probably reflecting the younger age structure of the Māori population. There were no substantial differences between Māori and non‐Māori in terms of either tumour grade or size, although Māori appeared to be somewhat more likely to have large tumours (>150 mm; 15% compared with 4% for non‐Māori) while the opposite was true for tumours with missing size data (35% and 49% respectively). The most common stage at presentation was stage IV (37% overall), and there was no statistically significant difference between Māori and non‐Māori for this variable ( p =0.65). Characteristics of 97 Māori and 92 non‐Māori liver cancer patients. Total (n=189) Māori (n=97) Non‐Māori (n=92) n % a n % b % c N % b % c P value d Sex Male Female 152 37 189 80% 20% 79 18 97 81% 19% 79% 21% 73 19 92 79% 21% 79% 21% 0.80 Age (years) 25–49 50–64 65–74 >75 Mean age (years) 41 65 48 35 189 19% 33% 27% 21% 61.9 29 38 20 10 97 30% 39% 21% 10% 58.6 ‐ ‐ ‐ ‐ ‐ 12 27 28 25 92 13% 29% 30% 27% 65.4 ‐ ‐ ‐ ‐ ‐ Tumour grade Poorly differentiated Moderately differentiated Well differentiated Missing 16 20 15 138 189 8% 10% 9% 73% 9 11 6 71 97 9% 11% 6% 73% 7% 13% 7% 73% 7 9 9 67 92 8% 10% 10% 73% 7% 10% 11% 72% 0.97 Tumour size (mm) <50 50–99 100–149 >150 Missing 49 26 19 16 79 189 26% 13% 11% 7% 44% 25 17 8 13 34 97 26% 18% 8% 13% 35% 26% 18% 8% 15% 33% 24 9 11 3 45 92 26% 10% 12% 3% 49% 26% 10% 12% 4% 49% 0.10 Stage Stage I Stage II Stage III Stage IV 29 35 57 68 189 15% 19% 29% 37% 16 16 32 33 97 16% 16% 33% 34% 18% 15% 35% 32% 13 19 25 35 92 14% 21% 27% 38% 15% 20% 26% 39% 0.65 Comorbid conditions Angina Hypertension Myocardial infarction Arrhythmia Valvular disease Congestive heart failure Cirrhosis CPD Diabetes Heavy alcohol use e Other primary cancer Renal disease 26 67 12 30 6 22 103 24 64 74 19 10 15% 33% 7% 17% 4% 12% 57% 13% 33% 40% 11% 6% 11 42 4 13 1 11 45 12 34 36 6 3 11% 43% 4% 13% 1% 11% 46% 12% 35% 37% 6% 3% 18% 51% 7% 18% 3% 17% 41% 17% 38% 35% 7% 5% 15 25 8 17 5 11 58 12 30 38 13 7 16% 27% 9% 18% 5% 12% 63% 13% 33% 41% 14% 8% 16% 25% 8% 18% 5% 11% 62% 13% 31% 39% 13% 7% 0.94 0.0007 0.54 0.60 0.19 0.31 0.002 0.36 0.36 0.40 0.28 0.56 Notes: Abbreviation: n, number; mm, millimetre; PVD, peripheral vasculardisease; CPD, chronic pulmonary disease Weighted Crude Age adjusted Chi square Current or previous heavy alcohol use noted Comorbidity was common overall with 57% of patients documented as having cirrhosis, 40% with documented heavy alcohol use, one‐third of patients with each of diabetes and hypertension, and more than 10% with each of cardiac arrhythmias, angina, congestive heart failure, chronic pulmonary diseases and other primary cancers. The prevalence of most of the observed comorbid conditions was similar for Māori and non‐Māori, except Māori were more likely to have hypertension (51% versus 25%) while non‐Māori were more likely to have cirrhosis recorded (62% versus 41%; both differences p <0.01). Table shows the weighted, crude and age‐adjusted proportions of all patients who received specific treatments. There were no substantial differences in specific treatment receipt between Māori and non‐Māori patients. Only 22% of all patients received curative surgery (resection or liver transplant), while 64% of the total study population were referred to palliative care. Receipt of treatment. Total (n=189) a Māori (n=97) Non‐Māori (n=92) n % n % b % c n % b % c P value d Definitive surgery Stage I Stage II Stage III Stage IV 15 19 8 1 43 37% 42% 18% 3% 7 12 5 0 24 29% 50% 21% 0% 22% 54% 23% 0% 8 7 3 1 19 42% 37% 16% 5% 30% 31% 8% 3% 0.56 Referred to Palliation 120 64% 61 63% 62% 59 64% 64% 0.96 Adjuvant treatment TACE Radiation PVE Palliative Chemo 38 12 6 2 21% 7% 4% 1% 18 5 2 0 19% 5% 2% 0% 23% 4% 1% 0% 20 7 4 2 22% 8% 4% 2% 21% 7% 5% 2% 0.65 0.66 0.20 0.08 Notes: Abbreviations: n, number; TACE, transcatheter arterial chemo‐embolization; PVE, Portal Vein Embolisation; Chemo, chemotherapy Weighted Crude Age adjusted Chi square Figure shows the age‐standardised prevalence of hepatitis for all patients. The age‐adjusted prevalence of hepatitis B among Māori patients (56%; 95% CI 45%‐67%) was more than double that of non‐Māori (27%; 95% CI 19%‐36%). The overall prevalence of hepatitis C was lower than that of hepatitis B, but was somewhat higher among non‐Māori (23%; 95% CI 15%‐31%) compared with Māori (14%; 95% CI 7%‐22%). Age standardised percentage of patients with viral hepatitis.* Table shows the weighted and crude proportions of patients recorded as having been under surveillance for HCC, by hepatitis status. A similar proportion of Māori and non‐Māori liver cancer patients with hepatitis B were under surveillance (37% and 39% respectively). Māori with hepatitis C were less likely to be on surveillance than non‐Māori with HCC (41% and 67% respectively; p =0.17). Nine non‐Māori patients were recorded as being under surveillance but having no viral hepatitis (29% of non‐Māori cohort on surveillance). In those patients with viral hepatitis who were on surveillance, 66% (35/53) were diagnosed with stage I or II cancers, whereas among those with viral hepatitis who were not on surveillance, 77% (54/70) were diagnosed at stage III or later. Surveillance status of those with Hepatitis B or C or no hepatitis. Total Māori Non‐Māori n % a n % b n % b P value c On surveillance Hepatitis B Hepatitis C No hepatitis 32 21 9 38% 59% 19% 23 7 0 37% 41% 0% 9 14 9 39% 67% 28% 0.95 0.17 0.007 Notes: Abbreviations: n, number on surveillance Weighted Crude Chi square Survival analysis suggested that Māori patients may have poorer survival compared with non‐Māori (age‐, sex‐ and stage‐adjusted hazard ratio: 1.36 [95% CI 0.96–1.92]). Discussion Māori were confirmed to have a considerably higher incidence of liver cancer than non‐Māori, and this difference showed no sign of decreasing from the 1980s to early 2000s – in absolute terms, the Māori:non‐Māori difference in incidence rates for males significantly increased. There were no significant differences in the tumour characteristics, including stage at diagnosis or treatment of Māori and non‐Māori patients with primary liver cancer. Documented heavy alcohol use was similarly common for both Māori and non‐Māori. There were, however, significant differences in the carriage of hepatitis B between ethnic groups. Māori with HCC were more than twice as likely to be hepatitis B‐positive compared to non‐Māori. The proportion of those with chronic hepatitis B who were under surveillance at diagnosis was similar between ethnic groups; however, less than 40% of both Māori and non‐Māori who developed HCC from chronic hepatitis B were on surveillance, and for Māori with hepatitis C the proportion was even lower. There was an indication of a survival difference for Māori compared to non‐Māori. The lack of difference in stage at diagnosis between Māori and non‐Māori is of interest. It is, in fact, consistent with some other recent studies. However, it is also important to note that there may have been minor differences in stage distribution, which we were not able to detect because of small numbers. The findings in relation to hepatitis are consistent with other NZ studies that have shown higher prevalence of hepatitis B infection and chronic hepatitis among Māori. Interestingly, the study by Blakely et al. based on a notes review of patients with HCC for the 1987–1994 period found a substantially greater difference in hepatitis B carrier prevalence between Māori liver cancer patients (77%) compared to Europeans (6%) than observed in the current investigation. It is possible that the rise of HCC due to hepatitis C, and possibly the reduction in hepatitis B due to vaccination programs (although it seems unlikely that this will yet be influencing HCC incidence), may mean that the stark differences in hepatitis B between ethnic groups have now lessened somewhat. It is also possible that the inclusion of non‐European ethnic groups in the non‐Māori group for this study, as well as the small numbers, affected the difference seen. There are three key strategies to decrease the incidence of HCC among Māori. First, vaccination to prevent hepatitis B infection is the main option for prevention. Vaccination is highly effective and is believed to confer immunity for at least 25 years. NZ has had a universal infant vaccination program in place since the late 1980s. The results of this vaccination program will manifest themselves over the coming decades in terms of a decrease in chronic hepatitis B infection and its sequelae, including HCC. It has, however, been noted that there is a reservoir of 30,000 potentially unidentified people with chronic hepatitis B that will hamper attempts to eradicate hepatitis B even with high rates of immunisation. Moreover, no vaccination is available for hepatitis C. The second strategy is screening for chronic hepatitis among those at higher risk. Screening involves the testing of at‐risk but asymptomatic individuals for the presence of hepatitis B surface antigen (HBsAg) or hepatitis C antibody and, if positive, HCV RNA. This confirms a diagnosis of chronic hepatitis B or hepatitis C infection respectively, and allows for surveillance and/or treatment with antivirals. The NZ hepatitis B screening programme from 1999 to 2002 screened 27% of Māori in the at‐risk population (i.e. those aged 15–40 years), well below its target coverage of 70%. Third, those who are screened and found have to chronic hepatitis B or C infection require surveillance or treatment. Periodic surveillance of patients with chronic hepatitis B, C or liver cirrhosis has been shown to result in an increase in early stage, asymptomatic cancers and a reduction in HCC mortality. Surveillance of patients at high risk of HCC most commonly involves use of two modalities: serum alpha feto‐protein levels (AFP) and abdominal ultrasound. Despite serum AFP lacking sensitivity (39% to 65% in previous studies) and ultrasound being user‐dependent and thought to lack specificity, surveillance by these two methods has been shown by randomised controlled trial to decrease HCC mortality by 37%. The authors of that trial found that 47% of the surveillance group underwent radical surgery compared with only 8% of the control group. In NZ, the Hepatitis Foundation maintains a follow‐up register for those with chronic hepatitis B and C to enable monitoring of liver function and surveillance for HCC. Those on surveillance in this study were diagnosed with earlier stage cancers, which is consistent with surveillance having a positive effect. However, this study was not designed to assess the effect of surveillance on mortality. Unfortunately, it was not possible to accurately ascertain the timing or intensity of screening from the clinical notes. The increasing incidence rates of liver cancer in both Māori and non‐Māori are in keeping with international trends of increasing rates, particularly in developed countries. Hepatitis C is an important cause of this increase in developed countries, with 50–70% of cases of HCC being attributed to hepatitis C in Japan and some European countries and, while hepatitis B remains more important in NZ, hepatitis C is clearly also becoming an important factor for liver cancer rates there. The rate of reported cirrhosis is surprisingly low given the high rates of referral for palliative care. However, this may be due to inconsistencies in documenting cirrhosis in the clinical notes. Heavy alcohol use was documented in 40% of cases and is likely to be an important factor in those cases without viral hepatitis. The survival analysis reported was based on small numbers and was under‐powered. Despite this, our best estimate is that Māori patients with HCC have 36% poorer survival than non‐Māori, which is consistent with documented survival disparities in other cancers in NZ. The Cancer Trends study links cancer registry with census data. It is in essence five consecutive cohort studies of the entire NZ population. The linkage of the two datasets overcomes numerator/denominator bias and misclassification of ethnicity, although the study is not completely without limitations. NZ is a small country, which can lead to statistical imprecision in stratifying results by ethnicity (more so for the smaller Asian and Pacific populations not included here). It was not possible to link all cancer registry records back to the census. To combat this, we corrected for linkage bias by using weights and are confident this has eliminated most bias due to misclassification of the outcome. The Cancer Registry Act of 1995 mandated the registration of all cancers. This may have caused a small artifactual rise in the liver cancer registrations following its introduction. Its effect is thought to be small and uniform throughout the ethnic groups. The notes review study is of a smaller cohort of patients. Because of this, the results may be affected by outliers and small differences will be difficult to detect with any certainty. We were able to collect data for all eligible patients, and this allowed an in‐depth review of the patients' tumour characteristics, co‐morbid conditions and interventions at multiple points along the treatment pathway. In summary, HCC remains an important health problem particularly for Māori men. Neonatal vaccination will over time decrease the prevalence of chronic hepatitis B infection in Māori and should reduce the disparities in the burden of liver cancer seen in this study. In addition, efforts to improve coverage of screening for hepatitis B and surveillance of those with chronic hepatitis should be a priority to immediately address the inequalities currently observed in liver cancer epidemiology. Acknowledgements The notes review component of this study was funded by the Health Research Council of NZ. The Cancer Trends component was funded by the Health Research Council of NZ and the Ministry of Health. The authors thank the Cancer, Comorbidity and Care (C3) research team, particularly Virginia Signal, who collected the notes review data, and James Stanley, who provided biostatistical support. We also thank June Atkinson for data analysis performed as part of the Cancer Trends study. Statistics New Zealand Security Statement: Access to the [Cancer Trends] data used in this study was provided by Statistics New Zealand in a secure environment designed to give effect to the confidentiality provisions of the Statistics Act 1975. The results in this study and any errors contained therein are those of the author, not Statistics New Zealand. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Australian and New Zealand Journal of Public Health Wiley

Incidence and management of hepatocellular carcinoma among Māori and non‐Māori New Zealanders

Download PDF
7 pages

Loading next page...
 
/lp/wiley/incidence-and-management-of-hepatocellular-carcinoma-among-m-ori-and-N0cuLW75T6

References (35)

Publisher
Wiley
Copyright
© 2013 Public Health Association of Australia
ISSN
1326-0200
eISSN
1753-6405
DOI
10.1111/1753-6405.12108
Publisher site
See Article on Publisher Site

Abstract

H epatocellular carcinoma (HCC or primary liver cancer) is one of the most common cancers worldwide. Annual global estimates range from 500,000 to 800,000 new cases. The Asia‐Pacific region is believed to contribute up to half of these, a number that reflects a high prevalence of major risk factors. Chronic hepatitis B and chronic hepatitis C account for 80% of primary liver cancer cases worldwide. Other risk factors include heavy alcohol consumption (particularly alcoholic cirrhosis) and dietary aflatoxin. Infant hepatitis B vaccination is decreasing the rate of chronic hepatitis B‐attributable HCC internationally. However, the number of cases of HCC globally is expected to treble by 2050 because of increasing chronic hepatitis C infection and non‐alcoholic fatty liver disease. In NZ, liver cancer is an important health issue, with 253 new registrations in 2009, and a high rate of mortality with 208 deaths recorded in the same period. Large ethnic disparities in incidence have previously been described, with high rates in Māori, Pacific and Asian populations. Blakely et al. found that Māori had seven times the rate of HCC compared with non‐Māori/Pacific/Chinese populations (in 1983–1994). There is insufficient evidence to say whether these disparities are changing over time. Mortality from HCC is also considerably higher for Māori. Furthermore, there is some evidence that survival for Māori with HCC may be lower than for non‐Māori. Survival from HCC is determined by the stage of cancer at diagnosis, with earlier stage cancers associated with improved outcomes. This is partly due to tumour biology, but is also related to whether aggressive therapy can be undertaken with curative intent (currently surgical resection and hepatic transplantation). Likelihood of survival is also determined by the presence and severity of associated hepatic cirrhosis, which is a risk factor for tumour development and reduces hepatic reserve, compromising the range of anti‐cancer therapies that can be employed. There is evidence in NZ from other cancer contexts that treatment differences may impact on disparities in outcomes, but this has not been studied for HCC. This study examined the incidence of HCC in Māori and non‐Māori New Zealanders over a 24‐year period (1981–2004) in more depth than previously undertaken, to investigate time trends in disparities. We reviewed the clinical notes of a sample of Māori and non‐Māori patients with HCC to determine differences in treatment pathways, comorbidity and care, which may affect disparities in outcomes. Methods There are four main ethnic groups in NZ. According to the 2006 census, 65% of the population identify as European, 14% as Māori, 7% as Pacific, and 9% as Asian, with around 11% identifying with other ethnic groups. This paper focuses on comparing those from the indigenous Māori population with those of other ethnicities. This paper draws upon two datasets. Incidence trends have been calculated using data from a national data linkage study (‘Cancer Trends’) and data on treatment pathways is drawn from clinical notes review. The methods of each component will be outlined separately. Cancer trends The New Zealand Cancer Registry (NZCR) is a population‐based register that collects information on all primary malignant tumours diagnosed in NZ, excluding squamous and basal cell skin cancers. Ethnicity data held by the NZCR is of variable quality, introducing numerator‐denominator bias in ethnic‐specific incidence rates that varies over time. For this reason, in studies investigating ethnic trends in cancer incidence, using census records linked to the NZCR is preferred. Five closed cohorts were created of the NZ usual resident population (25 years and older) on census night 1981, 1986, 1991, 1996, 2001, followed up for incident cancer(s) until the subsequent census or, in the case of 2001 cohort, until 31/12/2004. Cohorts were created using probabilistic record to anonymously link census and cancer register records within a geographic area (meshblock or census area unit) on sex, date of birth, ethnicity and country of birth. A modified total ethnicity approach was used. This includes individuals in all the ethnic groups that they identified with and leaves a residual non‐Māori/Pacific/Asian group, titled European/Other. For the purposes of this paper, Pacific and Asian populations have been excluded because of their small numbers (an average of about 15 Pacific and 10 Asian liver cancer cases per year over the study period). The linkage methods, processes and outputs for Cancer Trends have been detailed elsewhere. Briefly, 73.2% (1981–1986) to 81.7% (2001–2004) of eligible cancer registry records were linked to a census record, with 95.2% (1981–1986) to 96.9% (2001–2004) of these linked census‐cancer records estimated to be true links. Liver cancer, as it pertains to the Cancer Trends study, is defined as primary liver cancer (ICD code C22) and consists principally of hepatocellular carcinoma (the more common type), as well as cancer of the intrahepatic bile ducts (peripheral cholangiocarcinoma). Incidence rates and rate ratios (with 95% confidence intervals) were calculated after direct standardisation of the cohorts to the age structure of the WHO World Standard population. Statistical tests of trend were conducted for rates and of the log transformed rate ratios. All measures were also calculated for all five cohorts pooled. All analyses were conducted in SAS v9. Notes review Incident cases of HCC (ICD code C22.0) diagnosed between 01/01/2006 and 31/12/2008 were identified from the NZCR. Patients eligible for study inclusion were those aged 25 years or older at diagnosis, normally resident in NZ, who had not received a prior diagnosis of HCC and were diagnosed prior to death. Because more than 90% of Māori patients lived in the North Island, the notes review was limited to this area. All eligible Māori patients living in NZ's North Island along with a randomly sampled equal number of non‐Māori patients were included. Ethnicity information was drawn from the NZCR. Patients were classified as Māori if they had been identified as Māori on any previous health record. Clinical data were obtained from patients' medical records in both public and private hospitals. Data were recorded on a standardised study pro‐forma by a trained oncology nurse. Relevant data were double‐entered and discrepancies resolved. Data included: patients' symptoms on presentation; surveillance status (whether they were being surveyed for development of HCC); viral serology (HBV or HCV positive); tumour characteristics (tumour grade, size and stage at diagnosis [AJCC TNM staging system]); receipt of treatment (surgical, radiotherapy and chemotherapy); details of surgical intervention (type of surgeon, hospital and procedure as well as any pursuant complications); adjuvant care provided (reoperations, other operations); timing of treatment; the cancer treatment pathway (including referral, review, offering and receipt of treatments); and palliative care (referral and receipt). Mortality data were derived from the national Mortality Collection, with patients followed until 31/12/2010. Survey methods were used to create population estimates for the total NZ cohort of liver cancer patients over the study period. The final Māori and non‐Māori samples were weighted to the total eligible NZ Māori and non‐Māori liver cancer populations. Age‐standardised rates were calculated by direct standardisation using the total NZ cancer population. This adjusts for the differences in age structure between Māori and non‐Māori. Differences in trends between Māori and non‐Māori populations were calculated using Cochran‐Mantel‐Haenszel chi‐squared tests, stratified by age group. Survival trends were investigated using Cox proportional hazard regression models, adjusted for age (continuous variable), sex (male/female) and stage (I‐IV). Follow‐up time started at date of diagnosis and ended at date of death from liver cancer. Patients who died from other causes were censored at that date, and those who did not die were censored at 31/12/2010. All analyses were performed using SAS Version 9.2. This study was granted approval from the Multi‐Regional Ethics Committee (MEC 10/042/EXP). Results Cancer trends Figure shows the markedly higher incidence rates of HCC among Māori over all time periods and for both sexes. In males, the pooled (over time) annual rate for Māori of 19.6 per 100,000 (17.1–22.1) was nearly five times the rate for European/Other (RR=4.79; 95% CI 4.14–5.54). The pooled rate for Māori females was 6.1 (4.6–7.6) per 100,000 women and for European/Other females was 2.1 (1.9–2.3) per 100,000 women (RR=3.02; 95% CI 2.33–3.92). Standardised rates of liver cancer for 25+ year old patients per 100,000 per year. For both ethnic groups, the incidence of liver cancer increased between 1981 and 2004: by 90% for Māori males ( p for trend 0.03) and 79% for European/Other males ( p for trend <0.01), see Figure a. For European/Other females, incidence increased 29% ( p for trend 0.07) but Māori females' incidence rates were unstable over time (Figure b). The corresponding standardised rate ratios for Māori compared to European/Other ranged from 4.38 (2.79–6.88) in 1981–1986 to 5.08 (3.58–7.21) in 1986–1991 for males, with no evidence of increasing inequalities on a relative scale over the time period studied ( p for trend 0.90). Similarly, there was no evidence of increasing inequalities among females, with standardised ratios for Māori compared to European/Other varying between 4.69 (2.47–8.89) in 1981–1986 to 2.10 (1.03–4.31) in 1986–1991. However, in absolute terms the standardised rate difference did show a trend towards widening inequalities for males, increasing from 10 (5 to 15) to 19 (13 to 24) per 100,000 from 1981–1986 to 2001–2004 ( p for trend 0.04). No such trend in absolute inequalities was evident for females. Notes review There were 440 registrations for the ICD code C22.0 (HCC) in NZ during the study period (diagnosed 1/1/2006 to 31/12/2008). In the North Island, there were 117 Māori patients, 317 non‐Māori patients and six patients with missing ethnicity data (these were later merged with the non‐Māori cohort prior to random selection). After exclusion criteria were applied, there were 103 Māori diagnosed with liver cancer over the period studied in the North Island. This cohort was included along with a random sample of 103 from the 200 study‐eligible non‐Māori, also from the North Island. Following clinical notes review, 9% of patients were found ineligible for study inclusion. This comprised eight patients found to have been diagnosed outside of the study period, four patients with an element of cholangiocarcinoma (mixed tumours) to their diagnosis, three diagnosed and treated in Australia, one found to have metastatic liver disease, one with no evidence of a cancer diagnosis and one non‐resident. This resulted in a final sample of 97 Māori and 92 non‐Māori patients. The ethnicity division of the non‐Māori sample included 75% European ethnicity, 13% Pacific ethnicity and 11% Asian ethnicity. Table shows the weighted, crude and age‐adjusted characteristics of the 189 patients that comprise the final cohort. Around 80% of patients were men in both ethnic groups. Māori were diagnosed at a younger age than non‐Māori (mean age at diagnosis for Māori 58.6 years compared with 65.4 years for non‐Māori; p <0.001), probably reflecting the younger age structure of the Māori population. There were no substantial differences between Māori and non‐Māori in terms of either tumour grade or size, although Māori appeared to be somewhat more likely to have large tumours (>150 mm; 15% compared with 4% for non‐Māori) while the opposite was true for tumours with missing size data (35% and 49% respectively). The most common stage at presentation was stage IV (37% overall), and there was no statistically significant difference between Māori and non‐Māori for this variable ( p =0.65). Characteristics of 97 Māori and 92 non‐Māori liver cancer patients. Total (n=189) Māori (n=97) Non‐Māori (n=92) n % a n % b % c N % b % c P value d Sex Male Female 152 37 189 80% 20% 79 18 97 81% 19% 79% 21% 73 19 92 79% 21% 79% 21% 0.80 Age (years) 25–49 50–64 65–74 >75 Mean age (years) 41 65 48 35 189 19% 33% 27% 21% 61.9 29 38 20 10 97 30% 39% 21% 10% 58.6 ‐ ‐ ‐ ‐ ‐ 12 27 28 25 92 13% 29% 30% 27% 65.4 ‐ ‐ ‐ ‐ ‐ Tumour grade Poorly differentiated Moderately differentiated Well differentiated Missing 16 20 15 138 189 8% 10% 9% 73% 9 11 6 71 97 9% 11% 6% 73% 7% 13% 7% 73% 7 9 9 67 92 8% 10% 10% 73% 7% 10% 11% 72% 0.97 Tumour size (mm) <50 50–99 100–149 >150 Missing 49 26 19 16 79 189 26% 13% 11% 7% 44% 25 17 8 13 34 97 26% 18% 8% 13% 35% 26% 18% 8% 15% 33% 24 9 11 3 45 92 26% 10% 12% 3% 49% 26% 10% 12% 4% 49% 0.10 Stage Stage I Stage II Stage III Stage IV 29 35 57 68 189 15% 19% 29% 37% 16 16 32 33 97 16% 16% 33% 34% 18% 15% 35% 32% 13 19 25 35 92 14% 21% 27% 38% 15% 20% 26% 39% 0.65 Comorbid conditions Angina Hypertension Myocardial infarction Arrhythmia Valvular disease Congestive heart failure Cirrhosis CPD Diabetes Heavy alcohol use e Other primary cancer Renal disease 26 67 12 30 6 22 103 24 64 74 19 10 15% 33% 7% 17% 4% 12% 57% 13% 33% 40% 11% 6% 11 42 4 13 1 11 45 12 34 36 6 3 11% 43% 4% 13% 1% 11% 46% 12% 35% 37% 6% 3% 18% 51% 7% 18% 3% 17% 41% 17% 38% 35% 7% 5% 15 25 8 17 5 11 58 12 30 38 13 7 16% 27% 9% 18% 5% 12% 63% 13% 33% 41% 14% 8% 16% 25% 8% 18% 5% 11% 62% 13% 31% 39% 13% 7% 0.94 0.0007 0.54 0.60 0.19 0.31 0.002 0.36 0.36 0.40 0.28 0.56 Notes: Abbreviation: n, number; mm, millimetre; PVD, peripheral vasculardisease; CPD, chronic pulmonary disease Weighted Crude Age adjusted Chi square Current or previous heavy alcohol use noted Comorbidity was common overall with 57% of patients documented as having cirrhosis, 40% with documented heavy alcohol use, one‐third of patients with each of diabetes and hypertension, and more than 10% with each of cardiac arrhythmias, angina, congestive heart failure, chronic pulmonary diseases and other primary cancers. The prevalence of most of the observed comorbid conditions was similar for Māori and non‐Māori, except Māori were more likely to have hypertension (51% versus 25%) while non‐Māori were more likely to have cirrhosis recorded (62% versus 41%; both differences p <0.01). Table shows the weighted, crude and age‐adjusted proportions of all patients who received specific treatments. There were no substantial differences in specific treatment receipt between Māori and non‐Māori patients. Only 22% of all patients received curative surgery (resection or liver transplant), while 64% of the total study population were referred to palliative care. Receipt of treatment. Total (n=189) a Māori (n=97) Non‐Māori (n=92) n % n % b % c n % b % c P value d Definitive surgery Stage I Stage II Stage III Stage IV 15 19 8 1 43 37% 42% 18% 3% 7 12 5 0 24 29% 50% 21% 0% 22% 54% 23% 0% 8 7 3 1 19 42% 37% 16% 5% 30% 31% 8% 3% 0.56 Referred to Palliation 120 64% 61 63% 62% 59 64% 64% 0.96 Adjuvant treatment TACE Radiation PVE Palliative Chemo 38 12 6 2 21% 7% 4% 1% 18 5 2 0 19% 5% 2% 0% 23% 4% 1% 0% 20 7 4 2 22% 8% 4% 2% 21% 7% 5% 2% 0.65 0.66 0.20 0.08 Notes: Abbreviations: n, number; TACE, transcatheter arterial chemo‐embolization; PVE, Portal Vein Embolisation; Chemo, chemotherapy Weighted Crude Age adjusted Chi square Figure shows the age‐standardised prevalence of hepatitis for all patients. The age‐adjusted prevalence of hepatitis B among Māori patients (56%; 95% CI 45%‐67%) was more than double that of non‐Māori (27%; 95% CI 19%‐36%). The overall prevalence of hepatitis C was lower than that of hepatitis B, but was somewhat higher among non‐Māori (23%; 95% CI 15%‐31%) compared with Māori (14%; 95% CI 7%‐22%). Age standardised percentage of patients with viral hepatitis.* Table shows the weighted and crude proportions of patients recorded as having been under surveillance for HCC, by hepatitis status. A similar proportion of Māori and non‐Māori liver cancer patients with hepatitis B were under surveillance (37% and 39% respectively). Māori with hepatitis C were less likely to be on surveillance than non‐Māori with HCC (41% and 67% respectively; p =0.17). Nine non‐Māori patients were recorded as being under surveillance but having no viral hepatitis (29% of non‐Māori cohort on surveillance). In those patients with viral hepatitis who were on surveillance, 66% (35/53) were diagnosed with stage I or II cancers, whereas among those with viral hepatitis who were not on surveillance, 77% (54/70) were diagnosed at stage III or later. Surveillance status of those with Hepatitis B or C or no hepatitis. Total Māori Non‐Māori n % a n % b n % b P value c On surveillance Hepatitis B Hepatitis C No hepatitis 32 21 9 38% 59% 19% 23 7 0 37% 41% 0% 9 14 9 39% 67% 28% 0.95 0.17 0.007 Notes: Abbreviations: n, number on surveillance Weighted Crude Chi square Survival analysis suggested that Māori patients may have poorer survival compared with non‐Māori (age‐, sex‐ and stage‐adjusted hazard ratio: 1.36 [95% CI 0.96–1.92]). Discussion Māori were confirmed to have a considerably higher incidence of liver cancer than non‐Māori, and this difference showed no sign of decreasing from the 1980s to early 2000s – in absolute terms, the Māori:non‐Māori difference in incidence rates for males significantly increased. There were no significant differences in the tumour characteristics, including stage at diagnosis or treatment of Māori and non‐Māori patients with primary liver cancer. Documented heavy alcohol use was similarly common for both Māori and non‐Māori. There were, however, significant differences in the carriage of hepatitis B between ethnic groups. Māori with HCC were more than twice as likely to be hepatitis B‐positive compared to non‐Māori. The proportion of those with chronic hepatitis B who were under surveillance at diagnosis was similar between ethnic groups; however, less than 40% of both Māori and non‐Māori who developed HCC from chronic hepatitis B were on surveillance, and for Māori with hepatitis C the proportion was even lower. There was an indication of a survival difference for Māori compared to non‐Māori. The lack of difference in stage at diagnosis between Māori and non‐Māori is of interest. It is, in fact, consistent with some other recent studies. However, it is also important to note that there may have been minor differences in stage distribution, which we were not able to detect because of small numbers. The findings in relation to hepatitis are consistent with other NZ studies that have shown higher prevalence of hepatitis B infection and chronic hepatitis among Māori. Interestingly, the study by Blakely et al. based on a notes review of patients with HCC for the 1987–1994 period found a substantially greater difference in hepatitis B carrier prevalence between Māori liver cancer patients (77%) compared to Europeans (6%) than observed in the current investigation. It is possible that the rise of HCC due to hepatitis C, and possibly the reduction in hepatitis B due to vaccination programs (although it seems unlikely that this will yet be influencing HCC incidence), may mean that the stark differences in hepatitis B between ethnic groups have now lessened somewhat. It is also possible that the inclusion of non‐European ethnic groups in the non‐Māori group for this study, as well as the small numbers, affected the difference seen. There are three key strategies to decrease the incidence of HCC among Māori. First, vaccination to prevent hepatitis B infection is the main option for prevention. Vaccination is highly effective and is believed to confer immunity for at least 25 years. NZ has had a universal infant vaccination program in place since the late 1980s. The results of this vaccination program will manifest themselves over the coming decades in terms of a decrease in chronic hepatitis B infection and its sequelae, including HCC. It has, however, been noted that there is a reservoir of 30,000 potentially unidentified people with chronic hepatitis B that will hamper attempts to eradicate hepatitis B even with high rates of immunisation. Moreover, no vaccination is available for hepatitis C. The second strategy is screening for chronic hepatitis among those at higher risk. Screening involves the testing of at‐risk but asymptomatic individuals for the presence of hepatitis B surface antigen (HBsAg) or hepatitis C antibody and, if positive, HCV RNA. This confirms a diagnosis of chronic hepatitis B or hepatitis C infection respectively, and allows for surveillance and/or treatment with antivirals. The NZ hepatitis B screening programme from 1999 to 2002 screened 27% of Māori in the at‐risk population (i.e. those aged 15–40 years), well below its target coverage of 70%. Third, those who are screened and found have to chronic hepatitis B or C infection require surveillance or treatment. Periodic surveillance of patients with chronic hepatitis B, C or liver cirrhosis has been shown to result in an increase in early stage, asymptomatic cancers and a reduction in HCC mortality. Surveillance of patients at high risk of HCC most commonly involves use of two modalities: serum alpha feto‐protein levels (AFP) and abdominal ultrasound. Despite serum AFP lacking sensitivity (39% to 65% in previous studies) and ultrasound being user‐dependent and thought to lack specificity, surveillance by these two methods has been shown by randomised controlled trial to decrease HCC mortality by 37%. The authors of that trial found that 47% of the surveillance group underwent radical surgery compared with only 8% of the control group. In NZ, the Hepatitis Foundation maintains a follow‐up register for those with chronic hepatitis B and C to enable monitoring of liver function and surveillance for HCC. Those on surveillance in this study were diagnosed with earlier stage cancers, which is consistent with surveillance having a positive effect. However, this study was not designed to assess the effect of surveillance on mortality. Unfortunately, it was not possible to accurately ascertain the timing or intensity of screening from the clinical notes. The increasing incidence rates of liver cancer in both Māori and non‐Māori are in keeping with international trends of increasing rates, particularly in developed countries. Hepatitis C is an important cause of this increase in developed countries, with 50–70% of cases of HCC being attributed to hepatitis C in Japan and some European countries and, while hepatitis B remains more important in NZ, hepatitis C is clearly also becoming an important factor for liver cancer rates there. The rate of reported cirrhosis is surprisingly low given the high rates of referral for palliative care. However, this may be due to inconsistencies in documenting cirrhosis in the clinical notes. Heavy alcohol use was documented in 40% of cases and is likely to be an important factor in those cases without viral hepatitis. The survival analysis reported was based on small numbers and was under‐powered. Despite this, our best estimate is that Māori patients with HCC have 36% poorer survival than non‐Māori, which is consistent with documented survival disparities in other cancers in NZ. The Cancer Trends study links cancer registry with census data. It is in essence five consecutive cohort studies of the entire NZ population. The linkage of the two datasets overcomes numerator/denominator bias and misclassification of ethnicity, although the study is not completely without limitations. NZ is a small country, which can lead to statistical imprecision in stratifying results by ethnicity (more so for the smaller Asian and Pacific populations not included here). It was not possible to link all cancer registry records back to the census. To combat this, we corrected for linkage bias by using weights and are confident this has eliminated most bias due to misclassification of the outcome. The Cancer Registry Act of 1995 mandated the registration of all cancers. This may have caused a small artifactual rise in the liver cancer registrations following its introduction. Its effect is thought to be small and uniform throughout the ethnic groups. The notes review study is of a smaller cohort of patients. Because of this, the results may be affected by outliers and small differences will be difficult to detect with any certainty. We were able to collect data for all eligible patients, and this allowed an in‐depth review of the patients' tumour characteristics, co‐morbid conditions and interventions at multiple points along the treatment pathway. In summary, HCC remains an important health problem particularly for Māori men. Neonatal vaccination will over time decrease the prevalence of chronic hepatitis B infection in Māori and should reduce the disparities in the burden of liver cancer seen in this study. In addition, efforts to improve coverage of screening for hepatitis B and surveillance of those with chronic hepatitis should be a priority to immediately address the inequalities currently observed in liver cancer epidemiology. Acknowledgements The notes review component of this study was funded by the Health Research Council of NZ. The Cancer Trends component was funded by the Health Research Council of NZ and the Ministry of Health. The authors thank the Cancer, Comorbidity and Care (C3) research team, particularly Virginia Signal, who collected the notes review data, and James Stanley, who provided biostatistical support. We also thank June Atkinson for data analysis performed as part of the Cancer Trends study. Statistics New Zealand Security Statement: Access to the [Cancer Trends] data used in this study was provided by Statistics New Zealand in a secure environment designed to give effect to the confidentiality provisions of the Statistics Act 1975. The results in this study and any errors contained therein are those of the author, not Statistics New Zealand.

Journal

Australian and New Zealand Journal of Public Health – Wiley

Published: Dec 1, 2013

There are no references for this article.