Fax: (714) 456-7169
Changes in acquired immunodeficiency syndrome-related non-Hodgkin lymphoma in the era of highly active antiretroviral therapy
Incidence, presentation, treatment, and survival
Article first published online: 2 DEC 2005
Copyright © 2005 American Cancer Society
Volume 106, Issue 1, pages 128–135, 1 January 2006
How to Cite
Diamond, C., Taylor, T. H., Aboumrad, T. and Anton-Culver, H. (2006), Changes in acquired immunodeficiency syndrome-related non-Hodgkin lymphoma in the era of highly active antiretroviral therapy. Cancer, 106: 128–135. doi: 10.1002/cncr.21562
- Issue published online: 23 DEC 2005
- Article first published online: 2 DEC 2005
- Manuscript Accepted: 21 JUL 2005
- Manuscript Revised: 20 JUN 2005
- Manuscript Received: 17 MAY 2005
- National Cancer Institute. Grant Number: 1K07CA096480
- California Collaborative Treatment Group funded by the Universitywide Acquired Immunodeficiency Syndrome Research Program of the State of California. Grant Number: CC99-SD-003
- acquired immunodeficiency syndrome-related non-Hodgkin lymphoma;
- central nervous system;
- highly active antiretroviral therapy;
- histologic grade;
The authors sought to determine whether the availability of highly active antiretroviral therapy (HAART) coincided with changes in the epidemiology of acquired immunodeficiency syndrome (AIDS)-related non-Hodgkin lymphoma (NHL).
Cancer registry data from 1988–2000 were linked with AIDS registry data from 1981 to July 2003 for San Diego County to identify 537 AIDS-NHL patients. By using the total number of patients with AIDS who were alive as of July 1 annually as the AIDS population denominator, the average annual incidence of NHL was estimated among patients with AIDS for the pre-HAART period (1988–1995) and post-HAART period (1996–2000). The chi-square test was used to compare proportions, and a Cox proportional hazards model was used to compare survival between the pre-HAART and post-HAART periods.
The incidence of NHL decreased from 29.6 per 1000 person-years pre-HAART to 6.5 per 1000 person-years post-HAART. The proportion of patients who had NHL of central nervous system (CNS) origin decreased from 28% pre-HAART to 17% post-HAART. Among patients with systemic NHL, 54% received chemotherapy pre-HAART, and 72% received chemotherapy post-HAART. The percentage of intermediate-grade NHL increased from 33% pre-HAART to 49% post-HAART, and the percentage of high-grade NHL decreased from 38% to 19%, respectively. A diagnosis of human immunodeficiency virus infection preceding the NHL diagnosis and Stage IV NHL were associated with worse survival, whereas a diagnosis of NHL in the post-HAART period and chemotherapy were associated with better survival. The median survival was 4 months pre-HAART and 9 months post-HAART.
Since the introduction of HAART, there has been a decrease in the incidence of systemic and CNS NHL among patients with AIDS. Among patients with systemic, AIDS-related NHL, there has been decreased high-grade histology, increased use of chemotherapy, and improved survival. Cancer 2006. © 2005 American Cancer Society.
Highly active antiretroviral therapy (HAART) became available widely in the U.S. in 1996, and the proportion of human immunodeficiency virus (HIV)-infected outpatients who ever had received HAART increased from 37% in December 1996 to 71% by January 1998.1 Concurrent with the dissemination of HAART, acquired immunodeficiency syndrome (AIDS)-related mortality declined along with the incidence of opportunistic infections.2 Because individuals with AIDS have an increased risk of developing non-Hodgkin lymphoma (NHL),3 in the current study, we sought to determine whether the availability of HAART was associated with changes in the characteristics of AIDS-related NHL. We believe the current study is unique in that it is population-based rather than comprised of data from cohorts, clinical trials, or a single institution. By performing a linkage between the San Diego AIDS and cancer registries, we compared the incidence, presentation, treatment, and survival of AIDS-related NHL before and after the advent of HAART. We used data from > 500 patients with AIDS-related NHL that occurred over more than a decade.
MATERIALS AND METHODS
After we received approval from the University of California–Irvine Institutional Review Board (no. 2003–2940), we linked the AIDS and cancer registries for San Diego County in August 2003. For the linkage, cancer registry data from 1988 to 2000 were used, and AIDS registry data from 1981 through July 2003 were used. By using a computerized, matching algorithm that was enhanced by manual review of ambiguous cases, we identified 2055 patients who had both AIDS and a cancer diagnosis3–5; these patients were selected from 153,444 patients in the cancer registry and 11,867 patients in the AIDS registry. We identified 537 patients who had both AIDS and NHL from 5515 patients with NHL in the cancer registry.
Cancer Registry Methods
The San Diego/Imperial Organization for Cancer Control (SANDIOCC) is responsible for Region 7 of the California Cancer Registry, including operation of the population-based cancer registry for San Diego County. SANDIOCC receives information on all cancers diagnosed among San Diego County residents at any facility in California and adjacent states, including hospitals, outpatient clinics, pathology laboratories, the Coroner's office, and individual physicians' offices. In determining whether a cancer is reportable, the basic criterion is a diagnosis of cancer by a physician, even if the malignancy is not confirmed by pathology.4 However, a positive pathology report takes precedence over any other statement in a patient's chart.
SANDIOCC maintains a data base comprised of demographics, cancer type, extent of disease at diagnosis, treatment, and survival, as required by the State, and uses the International Classification of Diseases for Oncology, second edition,6 histologic coding. Registry staff conduct quality-control measures at all phases of data collection, through case-finding audits, reabstracting of cases, and consolidation of separate reports on the same tumor in a single patient. Staff periodically compare cancer registry incidence data against state mortality files that list cancer as the cause of death. During 1996–2000, only 1% of patients with NHL from San Diego County were identified on death certificates, suggesting nearly complete case ascertainment.7
AIDS Registry Methods
AIDS has been a reportable disease in California since 1981. The Centers for Disease Control established the original definition of AIDS in 1981 to encompass multiple opportunistic infections and malignancies (including NHL) and expanded the definition in 1993 (available at URL: http://www2.sdcounty.ca.gov/hhsa/documents/1993_revised_classification_system.pdf), adding a CD4 cell count of < 200/mm3 as an AIDS-defining condition.5, 8 When a physician makes an AIDS diagnosis, that physician is required by law to report the case to the local health department. The case report includes demographics, mode of HIV transmission, date and type of AIDS-defining illness, and the name, address, telephone number, and social security number of the patient. When the County of San Diego Health and Human Services Agency HIV/AIDS Epidemiology Unit investigates a reported AIDS case, an abstractor reviews the medical record to obtain the lowest, most recent, first, and serial CD4 counts. These values may be updated in the event of a duplicate report, but this is not uniform.
The County of San Diego Health and Human Services Agency HIV/AIDS Epidemiology Unit has longstanding relationships with many local HIV specialist providers as well as hospitals and clinics and conducts active surveillance. HIV infection became reportable in California in July 2002 and since that time, the HIV/AIDS Epidemiology Unit has received confirmed positive HIV antibody and all HIV viral load reports from laboratories that provide services to local healthcare providers, thereby enhancing the completeness of AIDS reporting. The HIV/AIDS Epidemiology Unit performs surveillance of death certificates and obtains updates of deaths from other jurisdictions to identify unreported and deceased AIDS patients.
We used SPSS statistical software (version 12.0; SPSS Inc., Chicago, IL), except as noted. We estimated the average annual incidence rates of NHL among patients with AIDS in San Diego County (with exact Poisson 95% confidence intervals9) for the pre-HAART and post-HAART periods. We defined 1988–1995 and 1996–2000 as the pre-HAART and post-HAART periods, respectively. We used the incident number of patients with AIDS-related NHL as the numerator and the sum of the patients with AIDS who were living as of July 1st annually as our AIDS population denominator (person-years at risk).
We used the Mann–Whitney U test to compare the median CD4 cell count and the time between HIV and NHL diagnoses in the pre-HAART and post-HAART periods. Using SAS software (version 9.0; SAS Inc., Cary, NC), we created a program to identify the CD4 cell count closest in time to the NHL diagnosis (within 1 yr before or 1 mo after NHL diagnosis) from the CD4 cell counts collected by the AIDS registry. In calculating the time between HIV and NHL diagnoses, we only included patients whose HIV diagnosis preceded their NHL diagnosis. We used the chi-square test to compare proportions between the pre-HAART and post-HAART periods, except when an expected cell count was < 5; then, we used the Fisher exact test. We defined CHOP as cyclophosphamide, doxorubicin, vincristine, and prednisone, allowing for substitution of doxorubicin with mitoxantrone or addition of etoposide, or bleomycin, or rituximab, or intrathecal methotrexate or cytarabine.10 We defined methotrexate-BACOD as any dosage of methotrexate with bleomycin, doxorubicin, cyclophosphamide, vincristine, and dexamethasone or prednisone with or without intrathecal cytarabine.11
We used the Kaplan–Meier method to determine the median survival of patients with systemic and CNS AIDS-related NHL. We defined NHL survival as the length of time in months between the NHL diagnosis and either death or the most recent follow-up. We then created a Cox proportional hazards model of survival. We performed univariate analyses for the following variables: age, gender, race/ethnicity, mode of HIV acquisition, period of diagnosis (pre-HAART vs. post-HAART), duration of HIV infection, receipt of chemotherapy, and NHL site, stage, and histologic grade. We then entered all variables that were statistically significant by univariate analysis into a forward conditional model to create a multivariate model of the characteristics associated with mortality from AIDS-related NHL.
Incidence of NHL among Patients with AIDS
Among 537 patients with AIDS-related NHL, 410 patients (76%) were diagnosed pre-HAART, and 127 patients (24%) were diagnosed post-HAART (Table 1). The peak number (n = 65 patients) occurred in 1994, and the nadir (n = 16 patients) occurred in the year 2000. There was a precipitous decline between 1995 and 1996, when the count decreased from 61 patients to 32 patients.
|Measure||Pre-HAART (1988–1995)||Post-HAART (1996–2000)|
|No. with incident NHL among patients with AIDS||410||127|
|No. with incident, systemic NHL among patients with AIDS||294||105|
|No. with incident CNS NHL among patients with AIDS||116||22|
|AIDS person-yrs at riska||13,874||19,535|
|Average annual incidence of NHL per 1000 patients with AIDS (95% CI))||29.6 (26.8–32.6)||6.5 (5.4–7.7)|
|Average annual incidence of systemic NHL per 1000 patients with AIDS (95% CI)||21.2 (18.8–23.8)||5.4 (4.4–6.5)|
|Average annual incidence of CNS NHL per 1000 patients with AIDS (95% CI)||8.4 (6.9–10.0)||1.1 (0.7–1.7)|
The average annual incidence of NHL (systemic and CNS) among patients living with AIDS decreased from 29.6 per 1000 person-years pre-HAART to 6.5 per 1000 person-years post-HAART (Table 1). The average annual incidence of systemic NHL decreased from 21.2 per 1000 person-years pre-HAART to 5.4 per 1000 person-years post-HAART. The average annual incidence of CNS NHL decreased from 8.4 per 1000 person-years pre-HAART to 1.1 per 1000 person-years post-HAART.
Characteristics of Patients with AIDS-Related NHL Pre-HAART and Post-HAART
The proportion of patients who had NHL of CNS origin decreased from 28% pre-HAART to 17% post-HAART (P < 0.01) (Table 1). Among 399 patients with systemic NHL, those diagnosed with NHL post-HAART were more likely to have a history of diagnosed HIV infection of at least 5 years' duration and were more likely to receive chemotherapy (Table 2). Excluding 32 patients with systemic NHL who received chemotherapy of unknown type, 141 patients (71%) received CHOP, 24 patients (12%) received methotrexate-BACOD, and 33 patients (17%) received other chemotherapy. These proportions did not change significantly between the pre-HAART and post-HAART periods (P < 0.62). The proportion of patients with Stage IV, systemic NHL also did not change (71% pre-HAART vs. 68% post-HAART; P < 0.49). The percentage of patients who had systemic NHL diagnosed on histologic or cytologic criteria, as opposed to radiologic, laboratory, clinical, or unknown criteria, increased from 94% (n = 277 patients) to 98% (n = 103 patients) post-HAART (P < 0.11). In contrast, the percentage of patients who had CNS NHL diagnosed on histologic or cytologic criteria decreased from 79% (n = 92 patients) to 68% (n = 15 patients) post-HAART (P < 0.27); the remaining patients predominantly received radiologic diagnoses. Among 138 patients who had CNS NHL, 71 patients (61%) received radiation therapy pre-HAART, 12 patients (55%) received radiation therapy post-HAART (P < 0.56), 10 patients (9%) received chemotherapy pre-HAART, and 4 patients (18%) received chemotherapy post-HAART (P < 0.24).
|Characteristic||No. of patients (%)||P value|
|Pre-HAART (1988–1995)||Post-HAART (1996–2000)|
|Total no. of patients||294||105|
|Duration of diagnosed HIV infection at time of NHL diagnosis||< 0.001a|
|HIV not diagnosed before NHL||65 (22)||9 (9)|
|> 0 days but < 1 yr||89 (30)||22 (21)|
|1–5 yrs||101 (34)||26 (25)|
|≥ 5 yrs||39 (13)||48 (46)|
|Histologic grade||< 0.001a|
|Intermediate grade||96 (33)||51 (49)|
|High grade||113 (38)||20 (19)|
|Unclassified and other||85 (29)||34 (32)|
|Received chemotherapy||155 (54)||75 (72)|
|Did not receive chemotherapy||133 (46)||29 (28)|
|Median time between HIV and NHL diagnosesc||21 mos||62 mos||< 0.001d|
|Median survival from time of NHL diagnosis||4 mos||9 mos||< 0.001e|
Comparing the histology in patients with systemic NHL who were diagnosed pre-HAART and post-HAART, the percentage of intermediate-grade NHL increased, and the percentage of high-grade NHL concomitantly decreased (Table 2). This was because of an increased frequency of intermediate-grade, diffuse, large cell lymphoma (histologic code 9680)6 from 21% of patients with systemic NHL pre-HAART to 44% post-HAART. Concurrently, there were decreases in high-grade, large cell, immunoblastic (diffuse) lymphoma (histologic code 9684) from 26% pre-HAART to 7% post-HAART. There was an increase in Burkitt lymphoma (histologic code 9687) from 4% pre-HAART to 9% post-HAART.
Among patients who were diagnosed with HIV before their NHL diagnoses, the median time between HIV and NHL diagnoses was shorter pre-HAART than post-HAART (Table 2). Among the 170 patients with systemic NHL who had known CD4 cell counts, the CD4 cell counts were 79/mm3 pre-HAART and 103/mm3 post-HAART (P < 0.29). Among the 58 patients with CNS NHL who had known CD4 cell counts, the median CD4 cell counts were 15/mm3 pre-HAART and 16/mm3 post-HAART (P < 0.98).
Survival of Patients with AIDS-Related NHL Pre-HAART and Post-HAART
Among patients with systemic NHL, the median survival was 4 months pre-HAART and 9 months post-HAART (P < 0.001) (Fig. 1). Among patients with CNS NHL, the median survival was 2 months pre-HAART and 1 month post-HAART (P < 0.39).
Characteristics Associated with Mortality in Patients with Systemic, AIDS-Related NHL
In univariate analyses of systemic, AIDS-related NHL, an HIV diagnosis before the NHL diagnosis, Stage IV NHL, high or unclassified histologic grade, an NHL diagnosis in the pre-HAART period, and the absence of chemotherapy were associated with decreased survival (Table 3). In multivariate analysis, an HIV diagnosis before the NHL diagnosis (P < 0.10 for HIV duration < 5 yrs), Stage IV NHL, an NHL diagnosis in the pre-HAART period, and the absence of chemotherapy were associated with worse survival. Except as noted, all P values for the multivariate analysis were < 0.01.
|Characteristic||No. of patients (%)||Crude relative hazard of death (95% CI)||Multivariate relative hazard of death (95% CI)|
|Age at NHL diagnosis|
|< 40 yrs||210 (53)||1.0 (reference)||—a|
|≥ 40 yrs||189 (47)||1.1 (0.9–1.4)||—a|
|Male||384 (96)||1.0 (reference)||—a|
|Female||15 (4)||0.8 (0.5–1.5)||—a|
|White||305 (76)||1.0 (reference)||—a|
|Latino||70 (18)||1.0 (0.7–1.3)||—a|
|Other||24 (6)||0.9 (0.6–1.4)||—a|
|Mode of HIV acquisition|
|Male-male intercourse||332 (83)||1.0 (reference)||—a|
|Injection drug use||21 (5)||1.2 (0.8–1.9)||—a|
|Male-male intercourse and injection drug use||26 (7)||1.1 (0.7–1.7)||—a|
|Other||20 (5)||0.9 (0.6–1.5)||—a|
|Period of diagnosis|
|Pre-HAART (1988–1995)||294 (74)||1.0 (reference)||1.0 (reference)|
|Post-HAART (1996–2000)||105 (26)||0.5 (0.4–0.7)||0.6 (0.5–0.8)|
|Duration of diagnosed HIV infection at the time of NHL diagnosis|
|HIV not diagnosed before NHL||74 (18)||1.0 (reference)||1.0 (reference)|
|> 0 days but < 1 yr||111 (28)||1.6 (1.1–2.1)||1.7 (1.3–2.4)|
|1–5 yrs||127 (32)||1.5 (1.1–2.1)||1.5 (1.1–2.1)|
|≥ 5 yrs||87 (22)||1.1 (0.8–1.5)||1.4 (0.9–2.0)|
|Site of lymphoma|
|Lymph node||284 (71)||1.0 (reference)||—a|
|Gastrointestinal||56 (14)||1.2 (0.9–1.6)||—a|
|Respiratory||24 (6)||1.2 (0.8–1.9)||—a|
|Other||35 (9)||0.8 (0.5–1.2)||—a|
|Stage of lymphomab|
|Stage IV||259 (70)||1.9 (1.4–2.4)||2.3 (1.8–3.0)|
|Stage I–III||109 (30)||1.0 (reference)||1.0 (reference)|
|Intermediate grade||147 (37)||1.0 (reference)||—a|
|High grade||133 (33)||1.4 (1.1–1.9)||—a|
|Unclassified and other||119 (30)||1.7 (1.3–2.2)||—a|
|Received chemotherapy||230 (59)||0.5 (0.4–0.6)||0.4 (0.3–0.5)|
|Did not receive chemotherapy||162 (41)||1.0 (reference)||1.0 (reference)|
We linked the San Diego AIDS and cancer registries to determine whether the availability of HAART coincided with changes in the epidemiology, presentation, treatment, and survival of patients with AIDS-related NHL. The incidence of AIDS-related NHL declined post-HAART along with the proportion of AIDS-related NHL of CNS origin. Among patients with systemic NHL, the percentage with high-grade histology decreased post-HAART. Patients with systemic NHL also were more likely to receive chemotherapy post-HAART. An HIV diagnosis before the NHL diagnosis, Stage IV NHL, an NHL diagnosis in the pre-HAART period, and the absence of chemotherapy were associated with worse survival for patients with systemic NHL.
The incidence of AIDS-related NHL decreased in San Diego County after the introduction of HAART. This is consistent with reports of decreased incidence of AIDS-related NHL in the post-HAART era from the U.K., France, Australia, and San Francisco, California.12–16 Our pre-HAART incidence of systemic NHL of 21 per 1000 person-years is slightly lower than the pre-HAART rate of 26 per 1000 person-years (95% confidence interval, 18–38 per 1000 person-yrs) reported by the Cascade collaboration17 for HIV-infected homosexual men with current CD4 cell counts < 100/mm3; however, our post-HAART incidence of systemic NHL of 5 per 1000 person-years is much lower than their post-HAART rate of 20 per 1000 person-years (95% confidence interval, 10–39 per 1000 person-yrs), most likely because our study population included patients who had CD4 cell counts > 100/mm3 and who therefore were at lower risk for developing NHL. Our post-HAART incidence was nearly identical to the post-HAART rate for systemic NHL (6–7 per 1000 person-yrs) described among patients with nadir CD4 cell counts < 144/mm3 from the Chelsea and Westminister HIV cohort.18
The decreased incidence of CNS lymphoma post-HAART reflects the shrinking population of AIDS patients in the U.S. with the extremely low CD4 cell counts typically observed in CNS NHL. Previous investigators have reported similar decreases in the frequency of CNS NHL post-HAART.13, 16, 19, 20 Although recent studies suggest that HAART-treated patients with CNS NHL live longer than in the past,21–23 registry patients with CNS NHL did not show improved survival post-HAART, possibly because of inconsistent antiretroviral use. Certainly, the low median CD4 cell counts among patients with CNS NHL in our study suggest a lack of treatment with HAART or the absence of a response to HAART.
The availability of HAART coincided with improved survival in AIDS patients with systemic NHL. This improvement may be related to better tolerance of chemotherapy and increased remission rates and/or to decreased occurrence of opportunistic infections among HAART-treated patients.24–27 The improved survival associated with chemotherapy in the current study may reflect a high initial performance status among patients who were treated with chemotherapy as well as a survival benefit from chemotherapy itself. The increased use of chemotherapy in the post-HAART period most likely is the result of the improved performance status and life expectancy of HAART-treated AIDS patients. Our post-HAART median survival duration of 9 months for patients with systemic NHL is shorter than expected, perhaps because our data included patients who did not receive HAART or chemotherapy. The median survival of patients with systemic NHL who receive current chemotherapy and HAART regimens likely would exceed our estimate.26, 27
An HIV diagnosis before the NHL diagnosis and Stage IV NHL were associated with worse survival in multivariate analysis. Although patients without a prior HIV diagnosis possibly would lack access to healthcare and certainly would not be receiving antiretroviral therapy (both characteristics associated with a poor prognosis), a preceding HIV diagnosis inherently would be associated with worse past performance status. Stage IV disease is a recognized sign of an unfavorable prognosis that has been included in previous studies of AIDS-related NHL outcomes and in the International Prognostic Index predictive model for aggressive NHL.11, 28, 29 Because histology was associated with period of diagnosis, receipt of chemotherapy, and disease stage, these associations may have caused histology to appear significant by univariate analysis.
Comparing the pre-HAART period with the post-HAART period, the percentage of patients with intermediate-grade NHL increased, whereas the percentage of high-grade NHL concurrently decreased. Because Burkitt lymphoma typically occurs in HIV patients who have higher CD4 cell counts, it would be reasonable to expect an increased incidence of Burkitt lymphoma post-HAART.30 However, the frequency of Burkitt lymphoma, which is a high-grade lymphoma, increased only slightly between the pre-HAART and post-HAART periods in our study. Levine et al. described an increase in diffuse large cell lymphoma in Los Angeles county with the advent of HAART.31 Little et al. proposed that the improved immune function associated with HAART results in fewer lymphomas of postgerminal-center origin, such as immunoblastic, diffuse large B-cell lymphoma; this may explain our finding of a decreased rate of this type of lymphoma in the post-HAART period.32, 33 We agree that improved immunologic control or decreased cytokine stimulation associated with effective HAART could cause a decreased frequency of herpesvirus-associated, immunoblastic NHL.34 Because the interpretation of NHL pathology can be subjective, and there is a high proportion of unclassified lymphomas,35 we remain cautious in our interpretation of this potentially important finding.
The study has additional limitations. The CD4 cell count data are incomplete. We have no information on individual patients' use of or response to HAART or other potential prognostic factors, such as performance status.28, 36 We do not know whether individuals living with AIDS emigrated from San Diego after enrollment in the registry, thus creating an artificially high AIDS population denominator and a spuriously low NHL incidence. We did not adjust our rates to exclude AIDS patients with prior NHL from the population at risk, and we did not adjust for race, gender, or HIV exposure category. However, because patients with AIDS-related NHL have a median survival of < 1 year, and our study population consisted predominantly of white men who have sex with men, we felt such adjustments were unnecessary. Finally, incomplete reporting to either the AIDS registry or the cancer registry always is a concern. Previous authors have estimated the completeness of this type of registry match for NHL at approximately 75%.37, 38 Because we do not have access to data on the patients with AIDS who did not match with patients with cancer, we cannot make such an estimate. However, the trends we describe likely would remain the same in the event of underreporting.
Although the incidence of AIDS-related NHL has decreased, NHL will remain an important cause of morbidity and mortality among patients with AIDS because of decreased rates of opportunistic infections.39 It would be interesting to determine whether HAART consistently influences the tumor histology of antiretroviral-treated patients who develop NHL and, if so, by what mechanisms. In addition, clinicians urgently require knowledge regarding how best to coordinate HAART and chemotherapy regimens. Therefore, further research will be necessary.
The authors are grateful to Michael Bursaw and Lorri Freitas (County of San Diego Health and Human Services Agency); to Phillip W. Virgo (Computer Sciences Corporation, Rockville, MD); and to Robert J. Biggar, Eric A. Engels, and James J. Goedert (Viral Epidemiology Branch, National Cancer Institute).
- 4California Cancer Registry Data Standard and Quality Control Unit. Cancer reporting in California: abstracting and coding procedures for hospitals. California Cancer Reporting System Standards. Sacramento: California Cancer Registry, 2003.
- 5Centers for Disease Control and Prevention. 1993 Revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. MMWR Morb Mortal Wkly Rep. 1992; 41: RR–17.
- 6PercyC, Van HoltenV, MuirC, editors. International classification of diseases for oncology, 2nd ed. Geneva: World Health Organization, 1990.
- 7Cancer incidence and mortality in San Diego County, 1996–2000. Irvine: Cancer Surveillance Program of Orange County/San Diego Imperial Organization for Cancer Control, Epidemiology Division, Department of Medicine, University of California–Irvine, 2003., , , et al.
- 8County of San Diego Health and Human Services Agency. County of San Diego HIV/AIDS Epidemiology Report 2004. San Diego: County of San Diego Health and Human Services Agency, 2004.
- 9Descriptive epidemiology. In: BreslowNE, DayNE, editors. Statistical methods in cancer research. Lyon: International Agency for Research on Cancer, 1994: 52, 63–64., , .
- 11National Institute of Allergy and Infectious Diseases AIDS Clinical Trials Group. Prognostic factors in the treatment of human immunodeficiency virus-associated non-Hodgkin's lymphoma: analysis of AIDS Clinical Trial Group Protocol 142—low-dose vs. standard-dose m-BACOD plus granulocyte-macrophage colony-stimulating factor. J Clin Oncol. 1998; 16: 3601–3606., , , , ,