The full content of Annals is available to subscribers

Subscribe/Learn More  >
Original Research |

A Prognostic Index for Systemic AIDS-Related Non-Hodgkin Lymphoma Treated in the Era of Highly Active Antiretroviral Therapy

Mark Bower, MA, PhD; Brian Gazzard, MD; Sundhiya Mandalia, PhD; Tom Newsom-Davis, MB BS; Christina Thirlwell, MB BS; Tony Dhillon, MB BS; Anne Marie Young, MB BS; Tom Powles, MD; Andrew Gaya, MB BS; Mark Nelson, MD; and Justin Stebbing, MA, PhD
[+] Article, Author, and Disclosure Information

From The Chelsea and Westminster Hospital, London, United Kingdom.

Potential Financial Conflicts of Interest: None disclosed.

Requests for Single Reprints: Mark Bower, MA, PhD, or Justin Stebbing, MA, PhD, The Chelsea and Westminster Hospital, 369 Fulham Road, London SW10 9NH, United Kingdom; e-mail, m.bower@imperial.ac.uk or j.stebbing@imperial.ac.uk.

Current Author Addresses: Drs. Bower, Gazzard, Mandalia, and Nelson: The Chelsea and Westminster Hospital, 369 Fulham Road, London SW10 9NH, United Kingdom.

Drs. Newsom-Davis, Dhillon, Young, and Powles: Charing Cross and Hammersmith Hospitals NHS Trust, London W6 8RF, United Kingdom.

Dr. Thirlwell: Cancer Research UK, Lincoln's Inn Fields, London WC2A 3PX, United Kingdom.

Dr. Gaya: Northwick Park Hospital, Watford Road, London HA1 3UJ, United Kingdom.

Dr. Stebbing: St. Bartholomew's Hospital, Bodley Scott Chemotherapy Unit, East Wing, West Smithfield, London EC1A 7BE, United Kingdom.

Author Contributions: Conception and design: M. Bower, B. Gazzard, S. Mandalia, M. Nelson, J. Stebbing.

Analysis and interpretation of the data: M. Bower, B. Gazzard, T. Newsom-Davis, C. Thirwell, T. Dhillon, J. Stebbing.

Drafting of the article: M. Bower, A. Gaya, J. Stebbing.

Critical revision of the article for important intellectual content: M. Bower, A.M. Young, T. Powles, A. Gaya, J. Stebbing.

Final approval of the article: M. Bower, B. Gazzard, C. Thirwell, T. Dhillon, A.M. Young, A. Gaya, J. Stebbing.

Provision of study materials or patients: M. Bower, A.M. Young, J. Stebbing.

Statistical expertise: M. Bower, S. Mandalia, J. Stebbing.

Obtaining of funding: M. Bower, B. Gazzard, J. Stebbing.

Administrative, technical, or logistic support: M. Bower, B. Gazzard, A.M. Young, J. Stebbing.

Collection and assembly of data: M. Bower, B. Gazzard, T. Dhillon, A.M. Young, J. Stebbing.

Ann Intern Med. 2005;143(4):265-273. doi:10.7326/0003-4819-143-4-200508160-00007
Text Size: A A A

The overall duration of survival was significantly greater for patients in whom non-Hodgkin lymphoma was diagnosed in the HAART era compared with those in whom the disease was diagnosed in the pre-HAART era (log-rank chi-square, 9.23; P = 0.002) (Figure 1). Among the 215 patients with AIDS-related non-Hodgkin lymphoma, the actuarial overall survival rate, including all causes of death, was 32% at 2 years (95% CI, 25% to 39%) and 26% at 5 years (CI, 19% to 33%).

First Page Preview

View Large
First page PDF preview


Grahic Jump Location
Figure 1.
Kaplan–Meier overall survival curve for 215 patients with systemic AIDS-related non-Hodgkin lymphoma (NHL) diagnosed in the era before (104 patients) and after (111 patients) highly active antiretroviral therapy (HAART).

All causes of death are included (log-rank chi-square, 9.23;  = 0.002).

Grahic Jump Location
Grahic Jump Location
Figure 2.
Product-limit survival plot for 111 patients with systemic AIDS-related non-Hodgkin lymphoma (NHL) diagnosed in the era of highly active antiretroviral therapy.

Patients are separated into whole-cohort quartiles of prognostic risk score. All causes of death are included.  < 0.001 (log-rank chi-square test).

Grahic Jump Location
Grahic Jump Location
Figure 3.
Receiver-operating characteristic curve showing sensitivity and false-positive error rate of mortality using the quartile cutoff values for the prognostic risk score, derived from the Cox proportional hazards regression coefficient.

Data from 111 patients are included. The sensitivity gives the degree of certainty that patients who fall in a particular prognostic risk score group will not die. The diagonal line displays ties, and the points on the curve refer to the sensitivity and 1 − specificity (the risk for false-positive results). 1 − specificity may also be referred to as the type I error rate. This demonstrates that the cutoffs established for risk scores are predictive of mortality.

Grahic Jump Location




Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).


Submit a Comment/Letter
To the Editor
Posted on August 23, 2005
Eiji Kusumi
The Fraternity Memorial Hospital
Conflict of Interest: None Declared

Bower et al. reported that CD4-positive cell counts at diagnosis and International Prognostic Index (IPI) except age were independent prognositc factors in post-highly active antiretroviral therapy patients with acquired immune deficiency syndrome-related lymphoma (ARL) (1). IPI at diagnosis is important in decision-making in malignant lymphoma in general (2). For example, since patients with high or high-intermediate risk IPI have poor prognosis, high-dose chemotherapy has been attempted (3, 4). The present study demonstrated important information in the management of ARL patients; however, information on the prognosis of ARL patients with low CD4 counts is insufficient. There are some possible reasons for the poor prognosis of these patients. The advanced immunodeficiency may render them at high risk of infections associated with chemotherapy, leading to high treatment-related mortality. Another possibility is that ARL with low CD4 - positive cell counts are resistant to chemotherapy. Managements would be different in these 2 cases; in the former, physicians may attempt to intensify infection management and to reduce the dose of chemotherapy, and in the latter, intensification of chemotherapy and use of different chemotherapeutic agents would be promising. Detailed description on the clinical courses and especially causes of death of ARL patients with low CD4 counts will provide many physicians with useful information. (205 words)

References 1. Bower M, Gazzard B, Mandalia S, et al. A prognostic index for systemic AIDS-related non-Hodgkin lymphoma treated in the era of highly active antiretroviral therapy. Ann Intern Med. 2005;143(4):265-73. 2. A predictive model for aggressive non-Hodgkin's lymphoma. The International Non-Hodgkin's Lymphoma Prognostic Factors Project. N Engl J Med. 1993;329(14):987-94. 3. Pettengell R, Radford JA, Morgenstern GR, et al. Survival benefit from high-dose therapy with autologous blood progenitor-cell transplantation in poor-prognosis non-Hodgkin's lymphoma. J Clin Oncol. 1996;14(2):586-92. 4. Gianni AM, Bregni M, Siena S, et al. High-dose chemotherapy and autologous bone marrow transplantation compared with MACOP-B in aggressive B-cell lymphoma. N Engl J Med. 1997;336(18):1290-7.

Conflict of Interest:

None declared

A Prognostic Index for Systemic AIDS-Related Non-Hodgkin's Lymphoma Treated in the Era of HAART
Posted on August 25, 2005
Regis A Vilchez
Department of Medicine, Section of Infectious Diseases, Rowan Regional Medical Center
Conflict of Interest: None Declared

To the Editor: Bower M, et al (1) examined the prognostic index for systemic AIDS-related non-Hodgkin's lymphoma (S-NHL) in the era of highly active antiretroviral therapy (HAART). The authors concluded that the international prognostic index and CD4 cell count predicted the prognosis of patients with AIDS- related S-NHL. Significant progress has been made in the management of this opportunistic disease during the last few years. However, outcomes still remain inferior compared to those achieved in HIV-negative individuals. These observations have led to the hypothesis of different tumor biology of lymphomas among HIV-infected patients as compared to - uninfected patients. Approximately two-thirds of AIDS-related S-NHLs are categorized as diffuse large B-cell type (DLBCL) with Burkitt lymphoma comprising 25% and other histologies a much smaller proportion (2). Recently, gene expression profilings of DLBCL among HIV-negative patients using nucleic acid microarrays (3) and immunohistochemistry (4) have shown that this single diagnostic category includes at least three gene expression subgroups, known as germinal center B-cell-like (GCB), activated B-cell-like and type 3. This last group is recognized to be a heterogeneous cluster that behaves in a manner similar to the ABC group. The GCB group has the best survival outcome. Although limited data exist on the phenotypic features of DLBCL in HIV-positive individuals, evidence suggests that tumor pathogenesis is heterogeneous (5-7) and that may be responsible for the outcome of this opportunistic complication of HIV infection (6,7). Indeed, a matched analysis of DLBCL cases from HIV- positive patients receiving HAART and HIV"“negative patients indicated that AIDS-related S-NHL had lower bcl-2 and higher CD10 expression, consistent with GCB profile and good prognosis (6). AIDS-related S-NHL cases were highly proliferative as compared to lymphomas among HIV-negative patients. High tumor proliferation did not correlate with poor outcome and may partially explain the high activity of chemotherapy among HIV-infected individuals receiving HAART (6). Another study of AIDS-related S-NHL showed that a non GCB profile was associated with a worse prognosis (7). Therefore, as data from different HIV cohorts continues to suggest that patients receiving HAART may develop lymphomas despite effective HIV suppression (2), it is important to develop a better understanding of the pathogenesis and biology of this opportunistic disease. In addition, biologic markers which identify the immunophenotypic features of lymphomas should be included in the prognosis, stratification and potential therapeutic decision strategies for AIDS-related S-NHL.

References: 1.) Bower M, Gazzard B, Mandalla S, et al. A prognostic index for systemic AIDS-related non-Hodgkin lymphoma treated in the era of highly active antiretroviral therapy. Ann Intern Med 2005;143:265-273. 2.) Navarro WH, Kaplan LD. AIDS-related lymphoproliferative disease. Blood 2005, in press. DOI 10.11.82/blood-2004-11-4278. 3.) Rosenwald A, Wright G, Chan WC, et al. The use of molecular profiling to predict survival after chemotherapy for diffuse large B-cell lymphoma. N Engl J Med 2002;346:1937-1947. 4.) Hans CP, Weisenburger DD, Greiner TC, et al. Confirmation of the molecular classification of diffuse large B-cell lymphoma by immunohistochemistry using a tissue microarray. Blood 2004;103:275-282. 5.) Vilchez RA, Lopez-Terrada D, Middleton JR, et al. Simian virus 40 tumor antigen expression and immunophenotypic profile of AIDS-related non- Hodgkin's lymphoma. Virology 2005, in press. DOI 10.1016/j.virol. 2005.06.053. 6.) Little RF, Pittaluga S, Grant N, et al. Highly effective treatment of acquired immunodeficiency syndrome-related lymphoma with dose adjusted EPOCH: impact of antiretroviral therapy suspension and tumor biology. Blood 2003;101:4653-4659. 7.) Hoffman C, Tiemann M, Schrader C, et al. AIDS-related B-cell lymphoma: correlation of prognosis with differentiation profiles assessed by immunophenotyping. Blood 2005, in press. DOI 10.1182/blood-2004-12-4631.

Conflict of Interest:

None declared

Prognostic modeling in AIDS-related lymphoma
Posted on October 21, 2005
Ewout W Steyerberg
Erasmus MC, Rotterdam, the Netherlands
Conflict of Interest: None Declared

Dear editor:

Recently a new prognostic index for a systemic AIDS-related lymphoma was proposed (1). The international prognostic index (IPI) and CD4 count where found to be the most important predictors of survival. This seems a reasonable conclusion, since CD4 count is a well-known predictor for patients with AIDS, while the IPI is well-known to predict prognosis of patients with lymphoma.

There are however a number of methodological problems in the underlying analyses. Although all of our major concerns were shared with the authors before publication, these were not incorporated in the paper. Others may hence accept the presented approach or elements of it. Specifically we previously referred the authors to a comprehensive checklist (2), which would have prevented most of the problems.

There are some technical errors in dealing with survival data with incomplete follow-up ("˜censoring'). Percentages of patients who died (Table 1, Table 4) are underestimates, and could better have been related to a particular point in time, e.g. two year after diagnosis (with Kaplan- Meier estimates). The ROC curve in figure 3 is invalid, since it does not account for follow-up time. More appropriate measures of discriminative ability are available, including the concordance statistic ("˜c'), which has a similar interpretation as the area under the ROC curve (3). Likelihood ratios are useful in a diagnostic context to indicate the extent to which a prior probability increases or decreases on the odds scale, but are confusing and invalid in a survival context with censoring. There are a number of other more minor mistakes, including use of the nonexistent entity "nonparametric data" (there are only nonparametric tests), and a hazard ratio of 1.01 for a reference category (reference categories of predictors are by definition 1.0).

Second, the authors categorized continuous prognostic variables. This results in a number of problems, related primarily to the arbitrariness of intervals and the implicit assumption of a flat prognostic impact within each interval of the predictor (2) (3). Adjusting for intervals of a variable and not for its entire spectrum results in residual confounding that causes other predictors to typically have exaggerated importance. We prefer flexible nonlinear approaches to model such prognostic effects, such as easily used regression splines (4).

Third, validation is essential unless patient samples are very large. As the authors correctly point out, the apparent performance of prognostic models is overestimated when assessed in small samples, compared to the performance observed in an independent sample. Bootstrapping is a valuable standard procedure for internal validation of model performance (3) (5) (6). The authors however bootstrapped the hazard ratio (HR) as estimated with Cox models, rather than an indicator of model performance such as the c statistic. This has never been suggested in any of the cited references to our methodological work. The mean HR is nearly unbiased, even in small samples, hence would not need to be bootstrapped. However, to take an example, high IPI had a Cox model HR of 4.9 [95% CI 1.5-15] and a bootstrapped HR of 6.7 [95% CI 6.0-7.3]. The authors seem to have calculated the bootstrap HR as mean(exp(coefficient)) instead of exp((mean(coefficient)), which explains the higher mean values by bootstrapping in Table 2. The bootstrap confidence intervals are too small by an order of magnitude, probably because they were based on standard errors (SE) rather the standard deviation (SD) (or because of a programming error). The SE will decrease towards zero with higher numbers of bootstraps, while the SD usually stabilizes after a few hundred bootstrap repetitions. Hence all presented bootstrap results should be ignored, and not be interpreted as demonstrating internal validity of the prognostic index.

Another issue is the presentation of the prognostic model. Prognostic weightings where derived by dividing coefficients by the lowest estimated coefficient. This approach has been followed before, but is statistically suboptimal. Preferably, multiplication is done by a constant, for example by 10, or in this case by 2. Such a multiplication results in a simple score which can be related to expected survival without grouping in quartiles (7).

Finally, the overall modeling strategy and interpretations are debatable. In total 215 patients were available with AIDS-related non- Hodgkin lymphoma, of whom 104 were diagnosed in the era before and 111 in the era after highly active anti-retroviral therapy (HAART) was available. The new index was based on only the latter 111 patients. This effectively assumes that interactions are present between prognostic factors and HAART, or, equivalently, that HAART has sub-group specific effects for all prognostic factors considered. This is highly unlikely, as supported by the reported finding that the IPI was of prognostic relevance in the pre- HAART era; the same is probably the case for CD4 count. A more natural starting point for model development would be to assume similar effects of prognostic factors across treatment strata (8). So, a better model might have been derived from considering the 215 patients, with the inclusion of a dummy variable indicating pre-HAART or post-HAART treatment era. As the authors indicate, additional predictors were identified as relevant among these 215 patients, including previous AIDS-defined illness and presence of Burkitt lymphoma. The fact that these factors were non-significant among the 111 HAART patients cannot be interpreted as evidence of lack of effects; only significant interaction tests could have supported this conclusion (8) (9).

In all, the authors may be qualitatively correct that IPI and CD4 count are important predictors in AIDS-related lymphoma. However the proposed index was not properly derived and not internally validated. A more valid index might have been obtained by modeling in the full data set, with shrinkage of regression coefficients to prevent too extreme estimates of survival differences (10). There is a need for more detailed recommendations on prognostic modeling, which may currently not be specific enough for authors submitting manuscripts to Annals of Internal Medicine (11).


Ewout W. Steyerberg, PhD Dept of Public Health Erasmus MC "“ University Medical Center Rotterdam Rotterdam, The Netherlands

Frank E. Harrell Jr, PhD Chair, Dept of Biostatistics Vanderbilt University School of Medicine Nashville, TN, USA


1. Bower M, Gazzard B, Mandalia S, et al. A prognostic index for systemic AIDS-related non-Hodgkin lymphoma treated in the era of highly active antiretroviral therapy. Ann Intern Med. 2005;143(4):265-73.

2. Harrell F, Byrne D. Statistical problems to document and to avoid. http://biostat.mc.vanderbilt.edu/twiki/bin/view/Main/ManuscriptChecklist. September 1, 2005.

3. Harrell FE. Regression modeling strategies: with applications to linear models, logistic regression, and survival analysis New York: Springer; 2001.

4. Harrell FE, Jr., Lee KL, Pollock BG. Regression models in clinical studies: determining relationships between predictors and response. J Natl Cancer Inst. 1988;80(15):1198-202.

5. Justice AC, Covinsky KE, Berlin JA. Assessing the generalizability of prognostic information. Ann Intern Med. 1999;130(6):515-24.

6. Steyerberg EW, Harrell FE, Jr., Borsboom GJ, Eijkemans MJ, Vergouwe Y, Habbema JD. Internal validation of predictive models: efficiency of some procedures for logistic regression analysis. J Clin Epidemiol. 2001;54(8):774-81.

7. Moons KG, Harrell FE, Steyerberg EW. Should scoring rules be based on odds ratios or regression coefficients? J Clin Epidemiol. 2002;55(10):1054-5.

8. Assmann SF, Pocock SJ, Enos LE, Kasten LE. Subgroup analysis and other (mis)uses of baseline data in clinical trials. Lancet. 2000;355(9209):1064-9.

9. Altman DG. Practical statistics for medical research. 1st ed London ; New York: Chapman and Hall; 1991.

10. Van Houwelingen JC, Le Cessie S. Predictive value of statistical models. Stat Med. 1990;9(11):1303-25.

11. Guidelines for preparing manuscripts. http://www.annals.org/shared/author_info.shtml#manuscriptpreparation. September 1, 2005.

Conflict of Interest:

None declared

In response
Posted on February 14, 2006
Justin Stebbing
No Affiliation
Conflict of Interest: None Declared

We appreciate the concerns of Drs Harrell Jr and Steyerberg and are grateful to them for their attention to detail. We address all of their points as follows:

In our paper (1), the numbers presented in Tables 1 and 4 are descriptive data not estimates. While we agree that Kaplan-Meier plots are appropriate, there are space limitations and we elected to present our data using hazard ratios (HRs) for individual factors considered to be of prognostic importance. These were derived using PHREG procedure in SAS where event time was defined as time since AIDS-related non-Hodgkin lymphoma (NHL) diagnosis to date of last follow-up or date of death. In Figure 1 of the paper, the upper (HAART-era) plot demonstrates the survival for this group. These patients unfortunately have a short lifespan as shown by a median survival of 0.71 years (95% CI 0.47 to 2.09), and the majority of deaths occur within the first year after diagnosis. If we only considered 2 years post diagnosis, as suggested, we would have very few patients in our sample, since the events (in this case death) occur earlier especially in those individuals with low CD4 counts, one of the findings of this study.

Prognostic risk scores were derived as stated in the methods section and constructed by dividing each £] coefficient in the final multivariable Cox¡¦s proportional hazards model with significant predictors by the lowest £] (to 2 decimal places). Using these pointƒn values, a risk score was assigned to each patient by adding up the points for each risk factor present. The prognostic risk score derived was then grouped into quartiles so that roughly equal number of patients fell in each of these categories. Unlike other prognostic scores, we did not round our values to the nearest integer, sacrificing clinical utility for statistical accuracy. By default the event time has been accounted for when constructing this since it is this that is used when preparing the hazard ratios. Cut-offs chosen for the prognostic risk score were further investigated using the ROC method which were used to demonstrate that using quartiles for the cut-offs, the scores that were derived from the proportional hazards multivariable model were appropriate. We acknowledge that the concordance statistic (¡¥c¡¦), is a useful and appropriate measure of discriminative ability, and has a similar interpretation to the area under the ROC curve; it measures 70.6% (95% CI 64.4 to 76.7%). We nevertheless used ungrouped prognostic risk score to derive area under the ROC curve as suggested and this was estimated as 83.3% (95% CI: 76.0% to 91.7%).

There is an important typographical error in Table 1: males should be the reference category with an HR of 1.0 (the correct HR for all reference categories as kindly mentioned) and the female HR measures 1.21 (95% CI 0.52 to 2.81). Use of the phrase non-parametric data has been used frequently in the literature (2) though the term would be more precisely worded as non-Gaussian distributed data. The categorization of continuous prognostic variables is a universal practice especially in prognostic modelling, and reasons for categorization often remain unstated (3-8). Nevertheless, we agree that any categorization must be justified. The prognostic variables we considered were categorized using quartiles of the CD4 count. While grouping these types of data using arbitrary cut-offs can be considered less robust, we have looked at the data that were grouped into 4 equally spaced categories and found that HRs were higher in individuals with CD4 counts that were less than 100 x 106 cells/L. Thus, this represented an appropriate clinically useful cut off. CD4 count measurement has high intra-patient variability and we considered changes on a small scale to be clinically irrelevant (9-11). The cut-off we chose for our work yielded almost equal numbers in each <100 and >100 groups which can be described as unbiased categorisation.

We are criticized for our technique of dividing each coefficient by a constant but Drs Harrell and Steyerberg propose an alternative of multiplying by a constant which is in fact mathematically identical. Drs Harrell and Steyerberg are however correct in pointing out that we calculated the bootstrap HR as mean(exp(coefficient)) instead of exp((mean(coefficient)) and we apologize for this error. This affected the confidence intervals for the bootstrap validation of the multivariable model. We have recalculated the confidence intervals and the editors have issued a correction for Table 2 that gives the recalculated intervals. The data have been resampled where the loge(HR) were averaged and the Efron¡¦s percentiles presented here use standard deviation as opposed to standard error. The prognostic index does not however change as this was calculated using the original data in the multivariable model.

Along with the peer reviewers of our paper, we believe that inclusion of only those individuals diagnosed with lymphoma in the HAART era is of clinical relevance for a number of reasons. HAART has dramatically reduced morbidity and mortality as a result of infection with HIV (12). As a consequence of this and also due to better treatments for cancer, there are dramatic improvements in AIDS-related NHL survival since the era of HAART (13, 14). Two comparisons of survival in AIDS-related NHL have shown an improvement in both response rate and overall survival (15, 16). Almost all major centers treat individuals with chemotherapy and concomitant (or occasionally sequential) HAART such that data from the 1980s and early 1990s is not applicable to current clinical practice (17, 18). This is reflected by the disappearance of prior AIDS-defining illnesses as prognostic factors in the HAART era (14), probably due to an improvement in the immune status of patients, as evident by an increase in the CD4 count at presentation during these periods.

This rigorous re-analysis of the data including use of Harrell¡¦s ¡¥c¡¦ statistic confirms the clinical utility of the prognostic model. The model demonstrates that for patients with AIDS-related NHL diagnosed in the era of HAART, the application of the International Prognostic Index remains useful. The addition of CD4 cell count provides further independent prognostic information. Patients who present with AIDS-related non-Hodgkin lymphoma and a low CD4 cell count have a poor prognosis and this information can be used to guide therapeutic options.


Mark Bower PhD FRCP

Brian Gazzard MD FRCP

Sundhiya Mandalia PhD

Justin Stebbing PhD MRCP

The Department of Oncology and HIV Medicine, The Chelsea and Westminster Hospital, 369 Fulham Road, London, SW10 9NH

Correspondence to j.stebbing@ic.ac.uk


We are grateful to Adam Sanitt for statistical advice.


1. Bower M, Gazzard B, Mandalia S, et al. A prognostic index for systemic AIDS-related non-Hodgkin lymphoma treated in the era of highly active antiretroviral therapy. Ann Intern Med. 2005;143(4):265-73. 2. Davies L, Wilkinson M, Bonner S, Calverley PM, Angus RM. "Hospital at home" versus hospital care in patients with exacerbations of chronic obstructive pulmonary disease: prospective randomised controlled trial. Bmj. 2000;321(7271):1265-8. 3. Landesman SH, Kalish LA, Burns DN, et al. Obstetrical factors and the transmission of human immunodeficiency virus type 1 from mother to child. The Women and Infants Transmission Study. N Engl J Med. 1996;334(25):1617- 23. 4. Preti HA, Cabanillas F, Talpaz M, Tucker SL, Seymour JF, Kurzrock R. Prognostic value of serum interleukin-6 in diffuse large-cell lymphoma. Ann Intern Med. 1997;127(3):186-94. 5. Kenchaiah S, Evans JC, Levy D, et al. Obesity and the risk of heart failure. N Engl J Med. 2002;347(5):305-13. 6. Bossuyt PM, Reitsma JB, Bruns DE, et al. The STARD statement for reporting studies of diagnostic accuracy: explanation and elaboration. Ann Intern Med. 2003;138(1):W1-12. 7. Anastos K, Barron Y, Cohen MH, et al. The prognostic importance of changes in CD4+ cell count and HIV-1 RNA level in women after initiating highly active antiretroviral therapy. Ann Intern Med. 2004;140(4):256-64. 8. O'Brien JM, Jr., Welsh CH, Fish RH, Ancukiewicz M, Kramer AM. Excess body weight is not independently associated with outcome in mechanically ventilated patients with acute lung injury. Ann Intern Med. 2004;140(5):338-45. 9. Aboulker JP, Autran B, Beldjord K, Touraine F, Debre P. Consistency of routine measurements of CD4+, CD8+ peripheral blood lymphocytes. J Immunol Methods. 1992;154(2):155-61. 10. Fei DT, Paxton H, Chen AB. Difficulties in precise quantitation of CD4+ T lymphocytes for clinical trials: a review. Biologicals. 1993;21(3):221-31. 11. Hughes MD, Stein DS, Gundacker HM, Valentine FT, Phair JP, Volberding PA. Within-subject variation in CD4 lymphocyte count in asymptomatic human immunodeficiency virus infection: implications for patient monitoring. J Infect Dis. 1994;169(1):28-36. 12. Palella FJ, Jr., Delaney KM, Moorman AC, et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. HIV Outpatient Study Investigators. N Engl J Med. 1998;338(13):853-60. 13. Matthews GV, Bower M, Mandalia S, Powles T, Nelson MR, Gazzard BG. Changes in acquired immunodeficiency syndrome-related lymphoma since the introduction of highly active antiretroviral therapy. Blood. 2000;96(8):2730-4. 14. Lim ST, Karim R, Tulpule A, Nathwani BN, Levine AM. Prognostic factors in HIV-related diffuse large-cell lymphoma: before versus after highly active antiretroviral therapy. J Clin Oncol. 2005;23(33):8477-82. 15. Navarro JT, Ribera JM, Oriol A, et al. Influence of highly active anti -retroviral therapy on response to treatment and survival in patients with acquired immunodeficiency syndrome-related non-Hodgkin's lymphoma treated with cyclophosphamide, hydroxydoxorubicin, vincristine and prednisone. Br J Haematol. 2001;112(4):909-15. 16. Vaccher E, Spina M, di Gennaro G, et al. Concomitant cyclophosphamide, doxorubicin, vincristine, and prednisone chemotherapy plus highly active antiretroviral therapy in patients with human immunodeficiency virus- related, non-Hodgkin lymphoma. Cancer. 2001;91(1):155-63. 17. Bower M, McCall-Peat N, Ryan N, et al. Protease inhibitors potentiate chemotherapy-induced neutropenia. Blood. 2004;104(9):2943-6. 18. Stebbing J, Marvin V, Bower M. The evidence-based treatment of AIDS- related non-Hodgkin's lymphoma. Cancer Treat Rev. 2004;30(3):249-53.

Conflict of Interest:

None declared

Submit a Comment/Letter

Summary for Patients

Estimating Outcome in Patients with HIV-Related Lymphoma

The summary below is from the full report titled, “A Prognostic Index for Systemic AIDS-Related Non-Hodgkin Lymphoma Treated in the Era of Highly Active Antiretroviral Therapy.” It is in the 16 August 2005 issue of Annals of Internal Medicine (volume 143, pages 265-273). The authors are M. Bower, B. Gazzard, S. Mandalia, T. Newsom-Davis, C. Thirlwell, T. Dhillon, A.M. Young, T. Powles, A. Gaya, M. Nelson, and J. Stebbing.


Clinical Slide Sets

Terms of Use

The In the Clinic® slide sets are owned and copyrighted by the American College of Physicians (ACP). All text, graphics, trademarks, and other intellectual property incorporated into the slide sets remain the sole and exclusive property of the ACP. The slide sets may be used only by the person who downloads or purchases them and only for the purpose of presenting them during not-for-profit educational activities. Users may incorporate the entire slide set or selected individual slides into their own teaching presentations but may not alter the content of the slides in any way or remove the ACP copyright notice. Users may make print copies for use as hand-outs for the audience the user is personally addressing but may not otherwise reproduce or distribute the slides by any means or media, including but not limited to sending them as e-mail attachments, posting them on Internet or Intranet sites, publishing them in meeting proceedings, or making them available for sale or distribution in any unauthorized form, without the express written permission of the ACP. Unauthorized use of the In the Clinic slide sets will constitute copyright infringement.


Buy Now for $32.00

to gain full access to the content and tools.

Want to Subscribe?

Learn more about subscription options

Related Articles
Topic Collections
PubMed Articles
Forgot your password?
Enter your username and email address. We'll send you a reminder to the email address on record.