| |
HCV, HIV & Non-Hodgins Lymphoma
|
| |
| |
"Non-Hodgkin's lymphoma and other nonhepatic malignancies in Swedish patients with hepatitis C virus infection"
Hepatology March 2005
See HIV Discusssion in Author Discussion-
"......NHL among the HIV-infected persons must be considered HIV-associated, but the HCV infection may also have affected this outcome....."
Ann-Sofi Duberg 1 *, Marie Nordström 2, Anna Törner 3, Olle Reichard 4, Reinhild Strauss 5, Ragnhild Janzon 3, Erik Bäck 1, Karl Ekdahl 3
1Department of Infectious Diseases, Örebro University Hospital, Örebro, Sweden
2Center of Hematology, Karolinska University Hospital, Stockholm, Sweden
3Department of Epidemiology, Swedish Institute for Infectious Disease Control, Solna, Sweden
4Department of Infectious Diseases, Karolinska University Hospital, Karolinska Institute, Stockholm, Sweden
5Department of Medical Epidemiology and Biostatistics, Karolinska Institute, Stockholm, Sweden
----Patients infected with hepatitis C virus (HCV) face a higher risk of developing non-Hodgkin's lymphoma and other nonhepatic malignancies, according to Swedish researchers.
"In Swedish hepatitis C patients, the risk of Non-Hodgkin's lymphoma and multiple myeloma were about 2 times higher than in the general population, though the risk is still not very high," Dr. Ann-Sofi Duberg from Orebro University Hospital told Reuters Health. "I think this knowledge is useful for doctors."
After identifying a cluster of four cases of non-Hodgkin's lymphoma (NHL) in young, otherwise healthy hepatitis C patients, Dr. Duberg and colleagues used data from 27,150 HCV-infected Swedes to evaluate the association between HCV infection and NHL, multiple myeloma (MM), thyroid cancer, chronic lymphatic lymphoma (CLL), acute lymphatic leukemia (ALL), and Hodgkin's lymphoma. Their results appear in the March issue of Hepatology.
The risk of NHL was 99% higher and the risk of MM was 154% higher among patients with HCV infection than among the general population in Sweden, the authors report. The increased NHL risk was especially significant in the group infected for more than 15 years.
In contrast, the report indicates, the risks of CLL and thyroid cancer were not significantly increased in HCV-infected patients.
The paucity of patients with ALL or Hodgkin's lymphoma (only 1 each) precluded any conclusions about their association with HCV infection, the researchers note.
"The hypothesis that HCV is involved in lymphomagenesis is supported by the fact that HCV-RNA can be detected in hematopoietic cells, particularly in B cells and monocytes," the investigators explain. "However, there is no convincing evidence of viral replication in these cells, and it is unclear how the virus induces B-cell proliferation."
As for the clinical implications, "Be aware of these malignancies in patients with hepatitis C and other symptoms consistent with lymphoma or myeloma, and, on the other hand, think of hepatitis C in patients with non-Hodgkin's lymphoma or multiple myeloma," Dr. Duberg advised.
"As chronic hepatitis C may last for decades without symptoms, it is often diagnosed when the patient is on a medical examination for some other reason," she noted.----
ABSTRACT
The aim of this study was to evaluate the association between hepatitis C virus (HCV) infection and non-Hodgkin's lymphoma (NHL), multiple myeloma (MM), thyroid cancer (TC), chronic lymphatic leukemia (CLL), acute lymphatic leukemia (ALL), and Hodgkin's lymphoma (HL). A Swedish cohort of 27,150 HCV-infected persons notified during 1990-2000 was included in the study. The database was linked to other national registers to calculate the observation time, expressed as person-years, and to identify all incident malignancies in the cohort. The patients were stratified according to assumed time of previous HCV infection.
The relative risk of malignancy was expressed as a standardized incidence ratio (SIR) - the observed number compared to the expected number. During 1990-2000 there were 50 NHL, 15 MM, 14 ALL, 8 TC, 6 CLL, and 4 HL diagnoses in the cohort. Altogether, 20 NHL, 7 MM, 5 TC, 4 CLL, 1 ALL, and 1 HL patient fulfilled the criteria to be included in the statistical analysis. The observation time was 122,272 person-years.
The risk of NHL and MM was significantly increased in the stratum with more than 15 years of infection (SIR 1.89 [standard incidence ratio] [95% CI, 1.10-3.03] and 2.54 [95% CI, 1.11-5.69], respectively). The association was not significant in TC or CLL.
In conclusion, we report the incidence of several malignancies in a nationwide cohort of HCV-infected persons. Although the delayed diagnosis of HCV probably has resulted in an underestimation of the risk, this study showed a significantly increased risk of NHL and MM.
HIV
We also looked at other underlying diagnoses that in other reports have been associated with the outcomes under study. According to the Hospital Discharge Register, 3 patients with NHL had undergone organ transplantation (1 liver, 2 renal). However, these patients were already excluded from analysis because of another cancer before NHL (n = 2) or NHL diagnosis preceding the HCV notification (n = 1). In the total HCV cohort, 149 patients had an HIV diagnosis. Six of these patients also had NHL, 4 of whom were in the group of 20 NHL patients included in the SIR analysis. No patients with CLL, ALL, HL, MM, or TC had an HIV diagnosis in the Hospital Discharge Register. When recalculating the SIR for the non-HIV NHL patients, the observation time was reduced to 121,859 person-years (35,927 and 85,932) and the SIR was 1.6 (95% CI, 0.91-2.59). In the two different strata, the SIRs were 1.86 (95% CI, 0.12-6.75) and 1.56 (95% CI, 0.86-2.62), respectively.
AUTHOR DISCUSSION
The role of HCV as an etiological factor or cofactor in the pathogenesis of NHL is controversial. Earlier studies have usually been case-control studies of NHL patients.[22] We used an opposite approach and studied the incidence of NHL, CLL, ALL, HL, MM, and TC in the entire cohort of notified HCV patients in Sweden. The results showed an increased risk for Swedish hepatitis C patients to develop B-cell NHL and MM.
The symptoms of acute hepatitis C are usually very mild, and the onset of the infection is rarely identified. The chronic phase may last for many decades without symptoms, and the infection could not be diagnosed until 1990. Since then, the infection is diagnosed when HCV screening is performed because of risk factors or coincidentally when the infected person is undergoing a medical examination for some other reason. This complicated the statistical analysis in the present study. The HCV diagnoses of as many as 14 of the 50 NHL patients and 6 of the 15 MM patients were notified less than 3 months before or after the cancer diagnosis; it cannot be excluded that the HCV infection was detected because of the malignancy, and these patients were therefore not included in the statistical analysis. However, the incubation period for hepatitis C is several months, and it seems likely that all of these patients were infected with HCV before the cancer diagnosis. According to our model on likely date of infection, all the patients in this group had been infected with HCV for more than 13 years, and the NHL or MM might be associated with HCV infection also in these cases. To disentangle this problem, we chose to give data both for the entire cohort and for a smaller cohort in which all of these patients had been excluded. Statistical analysis with SIR was possible only in the smaller cohort, which probably underestimated the association.
The risk for HCV-related malignancy might increase with time of HCV infection. The estimated median time of prior HCV infection was 16 years at time of notification. The eventual excess risk for developing malignancies that these patients suffered in the time they were infected with HCV but not yet notified remain undisclosed. For the stratum of patients infected less than 15 years, the relative risks for most of the studied malignancies were of the same magnitude as for the stratum with longer time of prior infection; however, the numbers are small, and no firm conclusions may be drawn. Because the date of infection is seldom known, this was estimated based on the probable route of transmission. Among HCV-positive intravenous drug users, 87% were born in 1950 or later, indicating that the spread of HCV in this population became more common from the late 1960s on. Non-A/non-B hepatitis (most cases of which were probably hepatitis C) existed in Sweden as early as the 1950s but became more common after IVDU became prevalent in the mid-1960s, and it has become an increasing problem.[21] In a study from Southern Sweden, stored frozen serum samples from patients with acute hepatitis between 1969 and 1972 were analyzed retrospectively, and among the intravenous drug users, 52% were positive for HCV antibodies.[26] In the 1990s, it was found that more than 90% of Swedish injection drug users were anti-HCV-positive by the age of 26 to 30 years,[27] and even occasional use of intravenous drugs was associated with a high risk of HCV infection.[28][29] This suggests that HCV infection is acquired early on in the life of intravenous drug users - during the last decade probably as early as the age of 15 to 25 years. These data formed the basis for the date of infection model for the IVDU group.
The majority of the NHL and MM patients were estimated to have been infected for more than 15 years, which is consistent with the theory that lymphomagenesis is a slow process and that NHL develops after long-lasting influence. The estimated date of infection often correlates with the age of the patient, and the cancer incidence increases with age, especially after the age of 50 years. In this study, 80% of the study population is 50 years or younger, and the expected number of malignancies is low, which might explain the small number of cancers. The median ages of the 50 patients with NHL and the 20 NHL patients included in the SIR analysis were 46 years (range, 9-87) and 44 years (range, 31-82), respectively (mean ages, 48 and 50 years, respectively) at date of NHL diagnosis. This is approximately 10 years lower than the median or mean age among NHL patients with hepatitis C in case-control studies that have found an association between lymphoma and HCV (mean age, 55-63 years).[7-9][13]
The hypothesis that HCV is involved in lymphomagenesis is supported by the fact that HCV-RNA can be detected in hematopoietic cells, particularly in B cells and monocytes. However, there is no convincing evidence of viral replication in these cells,[30] and it is unclear how the virus induces B-cell proliferation. It has been demonstrated that HCV-related lymphomas could be the result of HCV antigen stimulation of B cells.[18] Other indirect evidence for a possible role of HCV infection in the genesis of lymphoma is found in reports in which therapy against HCV is associated with regression of lymphoma. In one study of 9 patients with HCV infection and splenic marginal zone lymphoma, all 9 patients had a regression of the lymphoma as they cleared the virus when treated with interferon alfa with or without ribavirin.[31]
Most other studies have analyzed the prevalence of HCV markers among patients with B-cell NHL. Studies from Italy, Japan, and North America have found an association,[8][10][11] but a large case-control study from France[13] and other studies have not.[12] A geographical difference has been suggested, maybe because of the lower HCV prevalence in some areas or other cofactors such as HCV genotype. In an Italian study, the prevalence of HCV genotype 2ac was significantly increased among patients with B-cell NHL (48.4%) compared with the control group (9%),[32] but in other studies no such correlation has been found.[8] Sweden has a low prevalence of HCV; however, in this study we investigated the incidence of lymphoproliferative disorders among the entire cohort of notified HCV patients, and the result should not be affected by the low HCV prevalence in Swedish society. In Sweden, there is a different HCV genotype distribution compared with most other parts of the world with a lower frequency of HCV genotype 1 (especially 1b). In a Swedish study, genotypes 1a (35%) and 3 (31%) were the most common genotypes, followed by genotype 2 (17%) and 1b (6%).[33] The genotype distribution among the patients in this study is not known.
Because HCV is a blood-borne virus, HCV-infected persons have been at risk of catching other, yet unknown blood-borne infections that might play a role in the development of malignant tumors.
HIV infection is a well-recognized risk factor for malignant lymphoma. Lymphoma is more common among individuals with a severe immune deficiency and is considered an AIDS-defining event. The incidence of NHL among HIV-infected patients has decreased significantly since the introduction of highly active antiretroviral therapy in 1996.[14] The exact incidence of HIV-related lymphoma in Sweden is not known. In Sweden, HIV is a notifiable disease (via anonymous notification). Between 1985 and 2000, 5,368 HIV-infected individuals were reported; of these, 805 were intravenous drug users, many of them probably coinfected with HCV. Approximately 3,000 of the notified HIV-infected persons are still alive.[34] In a study from southern Sweden, 2.1% of intravenous drug users were HIV-positive.[27]
In this study, 149 of 27,150 HCV patients had an HIV diagnosis according to the Hospital Discharge Register. This is likely an underestimation, because many HIV-infected patients have not been hospitalized. However, all but 1 of the patients with NHL had been hospitalized, and 6 of the 50 patients with HCV and NHL also had an HIV diagnosis. Some studies have investigated the role of HCV coinfection in HIV-infected individuals but found no effect on the risk for NHL.[35][36] In this study, we present the results with and without the HIV-infected NHL patients. NHL among the HIV-infected persons must be considered HIV-associated, but the HCV infection may also have affected this outcome. This is also the case in transplanted patients, who may have an increased risk to develop NHL,[15] with a reported higher risk of this complication in HCV-positive liver transplant recipients.[37] In our study, the 2 transplanted patients who had a hepatitis C diagnosis before NHL and could have been included in the statistical analysis were nevertheless excluded because another cancer had preceded the NHL.
Among Swedish blood donors, the prevalence of HCV antibodies has been 0.23% to 0.48%.[38] The prevalence in the general population in Sweden is not known. In this study we included every individual in Sweden who was notified with an HCV diagnosis since 1990 - when the diagnostic methods for HCV became available - until the end of 2000. Most HCV diagnoses are notified, but the extent of missed notifications is not known.
The HCV diagnosis is often based on the detection of HCV antibodies (using second- or third-generation enzyme-linked immunosorbent assay) and is confirmed with a Recombinant Immuno Blot Assay (better known as RIBA), which has been widely used in Swedish laboratories. Detection of HCV RNA via polymerase chain reaction has became more common, and during the late 1990s, most HCV infections were confirmed using HCV RNA tests. Because HCV diagnosis for many patients is based on the detection of HCV antibodies, some of the patients in this study might be persons with a resolved HCV infection either as a part of the natural history of hepatitis C or as a result of treatment of hepatitis C. On the other hand, several case-control studies on HCV and lymphoma have found that an HCV antibody test without HCV RNA analysis may underestimate the true rate of HCV infection in patients with NHL.[7][8][11]
In conclusion, this study examined the incidence of NHL, MM, CLL, ALL, HL, and thyroid cancer in a nationwide cohort of HCV-infected persons. Although the delayed diagnosis of hepatitis C most probably has made us underestimate the risk, this study showed that the risks of B-cell NHL and MM (but not CLL, ALL, HL, and TC) were significantly increased. Furthermore, the relatively late spread of HCV in Sweden, with 80% of the HCV-infected individuals born in 1950 or later, might have resulted in an underestimation of the overall long-term risk for HCV-related malignancies. This indicates that further studies - perhaps on a smaller but better-characterized cohort of Swedish patients - would be of great interest.
Article Text
The characterization of hepatitis C virus (HCV) in 1989[1][2] and the subsequent development of sensitive and specific diagnostic assays have revealed that chronic HCV infection is a global problem, affecting more than 170 million individuals worldwide.[3] Although the majority of infected persons have few or no clinical symptoms, individuals with chronic HCV infection have an increased risk of developing cirrhosis and hepatocellular carcinoma.[4][5] Furthermore, an association between HCV and essential mixed cryoglobulinaemia has been established.[6] The role of HCV as an etiological factor or cofactor in the pathogenesis of other lymphoproliferative disorders is, however, controversial. In some studies, the prevalence of HCV infection was significantly increased among patients with B-cell non-Hodgkin's lymphoma (NHL). This was especially the case in studies performed in Italy, but was also seen in other countries such as Japan and the United States.[7-11] However, this association has not been demonstrated in studies from Northern Europe, and regional differences have been suggested.[12][13] Moreover, HCV has been associated with other nonhepatic neoplasms such as multiple myeloma (MM) and thyroid cancer (TC).[9]
The mechanism through which HCV may cause lymphoproliferative disorders is not fully known. In general, risk factors for malignant lymphomas are mainly correlated to an altered or depressed immune system, such as human immunodeficiency virus (HIV) infection, immunosuppressive medication, severe inborn immunological deficiencies, or rheumatic diseases.[14-16] HCV has been suggested to be both hepatotropic and lymphotropic.[17][18]
The incidence of malignant NHL has increased in many countries in recent years.[19] In Sweden the average annual increase between 1981 and 2000 was 1.2% for men and 1.5% for women.[20] The majority of Swedish HCV patients are born after 1950, indicating that the large-scale spread of HCV started in the late 1960s and 1970s.[21] We may now be seeing an increasing number of patients with complications due to long-lasting HCV infection.
A cluster of four incident cases of NHL in young, otherwise healthy hepatitis C patients in a cohort of 554 HCV-infected persons raised the question of an association between HCV and NHL in Sweden. The aim of this study was to evaluate the association between HCV infection and NHL in a Northern European country where this association had not been previously described. Furthermore, we investigated the association between HCV and other nonhepatic malignancies in which an association had been described earlier or with a theoretic background consistent with an association - that is, MM, TC, chronic lymphatic leukemia (CLL), acute lymphatic leukemia (ALL), and Hodgkin's lymphoma (HL).
In contrast to previous case-control studies involving patients with malignancies who were tested for HCV,[9][22] we studied the entire cohort of notified HCV patients in Sweden by linking different nationwide databases.
Study Population
In Sweden, HCV infection (both acute and chronic) has been, by law, a notifiable disease since 1990, when diagnostic methods for HCV became available. Notifications are made in duplicate to the Swedish Institute for Infectious Disease Control (SMI); one from the clinician and one from the laboratory having diagnosed the infection. At the SMI, the notifications are merged using a unique 10-digit personal identification number that is issued to all Swedish residents[23] and is used in all health care contacts. The clinical notifications include information of epidemiological relevance, such as route of transmission.
From 1990 to 2000, 30,180 notifications of hepatitis C were made to the SMI. After excluding duplicate notifications and individuals with an incorrect personal identification number, 27,150 HCV patients were included in the cohort. For each patient we retrieved information on the personal identification number, sex, date of HCV notification, and presumed route of transmission.
Linkage to Other Registers
By using the personal identification number, the database with the HCV patients could be linked to other national registers. Information on dates of immigration, emigration, and deaths were obtained from the Swedish Population Register held at Statistics Sweden.
All cases of cancer (in this study, the term cancer includes all of the malignant diseases under study) in Sweden are reported to the Swedish Cancer Registry both by clinicians and pathologists/cytologists, and it has been estimated that more than 95% of all tumors are reported.[24] The cancers are coded according to the seventh revision of the International Classification of Diseases (ICD-7). We used the Cancer Registry to identify all incident cancers diagnosed in the cohort between 1990 and 2000, with special interest in all episodes of malignant NHL (ICD-7 codes 200.1, 200.2, 200.3, 202.1, 202.2), CLL (ICD-7 code 204.1), ALL (ICD-7 code 204.0), HL (ICD-7 code 201), MM (ICD-7 code 203.0), and TC (ICD-7 code 194.0).
The Swedish Hospital Discharge Register compiles data on discharge diagnoses from every hospital in Sweden; up to eight medical conditions are filed for each in-hospital episode. The discharge diagnoses were coded according to the ninth revision of the ICD (ICD-9) until 1997; since then they have been coded according to the tenth revision (ICD-10). Validations of the Hospital Discharge Register have demonstrated a correct ICD code at the four-digit level in 86% of all principle discharge diagnoses.[25] We obtained information regarding organ transplantation (ICD-9 codes V42A-B and V42G-H and ICD-10 codes Z94.0-Z94.4) and any HIV-associated diagnosis (ICD-9 codes 079J, 279K, and V02J and ICD-10 codes B20-B24 and Z21) from the Hospital Discharge Register for those patients that had been hospitalized.
|
|
| |
| |
|
|
|
