Role of serum free light chains in predicting HIV-associated
non-Hodgkin lymphoma and Hodgkin’s lymphoma and its
correlation with antiretroviral therapy
Michele Bibas,1* Maria Paola Trotta,1Alessandro Cozzi-Lepri,2,3Patrizia Lorenzini,1Carmela Pinnetti,1
Giuliano Rizzardini,4Gioacchino Angarano,5Pietro Caramello,6Laura Sighinolfi,7
Claudio Maria Mastroianni,8Giovanni Mazzarello,9Antonino Di Caro,1Cristina Di Giacomo,1
Antonella d’Arminio Monforte,10and Andrea Antinori1; for the ICONA Foundation Study Group
A nested case-control study was performed within the Italian cohort of naı ¨ve to antiretroviral human immuno-
deficiency virus (HIV) patients (ICONA) cohort to evaluate the role of serum free light chains (sFLC) in predict-
ing non-Hodgkin’s lymphoma (NHL) and Hodgkin lymphoma (HL) in HIV-infected individuals. Of 6513 partici-
pants, 86 patients developed lymphoma and 46 of these (NHL, 30; HL, 16) were included in this analysis hav-
ing stored prediagnostic blood. A total of 46 serum case samples matched 1:1 to lymphoma-free serum
control samples were assayed for j and k sFLC levels and compared by using conditional logistic regression.
Because the polyclonal nature of free light chains (FLCs) was the focus of our study, we introduced the k 1 k
sum as the measurement of choice and as the primary variable studied. j 1 k sFLC values were significantly
higher in patient with lymphoma than in controls, especially when considering samples stored 0–2-year pe-
riod before the lymphoma diagnosis. In the multivariable analysis, the elevation of sFLC predicted the risk of
lymphoma independently of CD4 count, (odd ratio of 16.85 for k 1 k sFLC >2-fold upper normal limit (UNL) vs.
normal value). A significant reduction in the risk of lymphoma (odd ratio of 0.07 in model with k 1 k sFLC)
was found in people with low sFLC and undetectable HIV viremia lasting more than 6 months. Our analysis
indicates that an elevated polyclonal sFLC is a strong and sensitive predictor of the risk of developing lym-
phomas, and it is an easy to measure biomarker that merits consideration for introduction in routine clinical
practice in people with HIV. Am. J. Hematol. 87:749–753, 2012.
C 2012 Wiley Periodicals, Inc.
Malignancies and in particular lymphomas remain an impor-
tant cause of morbidity and mortality among patients with
human immunodeficiency virus (HIV) infection, although a sub-
stantial improvement in survival has been reported following
the introduction of highly active antiretroviral therapy (HAART)
[1,2]. Data from cohort studies showed an overall reduction in
the incidence of non-Hodgkin’s lymphoma (NHL), but not in
Hodgkin lymphoma (HL) [3–6]. In this context, there is an
urgent need for blood-based noninvasive tests that may assist
in the lymphoma risk assessment of HIV-infected individuals.
Currently, only age, CD4 cell count, HIV viral load, and use of
HAART have been shown to play a role as risk factors for
malignancies in HIV-infected individuals [7,8].
Great interest has been devoted recently to the role of B-
cell dysfunction in the etiology of AIDS-related NHL through
direct measurement of serum-based markers of B-cell stim-
ulation [8–12]. Landgren et al. have found a strong associa-
tion of elevated polyclonal serum free light chains (sFLC)
with the risk of AIDS-related NHL . This study referred
mainly to the pre-HAART era, and most patients were not
receiving HAART; consequently, little information of the role
of HAART on this marker was presented, and no HL
patients were included. Recently, the elevated sFLC levels
were associated with poor outcome in non-HIV-infected
patients with diffuse large B-cell lymphomas (DLBCL), HL,
and in chronic lymphocytic leukemia [14–16].
The aim of this study was then to evaluate the associa-
tion of sFLC levels with the risk of both NHL and HL in
HIV-infected patients and its correlation with combination
Patients and Methods
the Italian cohort of naı ¨ve to antiretroviral HIV patients (ICONA). For a
We performed a case-control study nested within
1National Institute for Infectious Diseases ‘‘Lazzaro Spallanzani’’ IRCCS, Clinical Department, Rome, Italy;2University College London, London, Department
of Infection & Population Health, Division of Population Health, United Kingdom;3Division of Biostatistics, University of Minnesota, Minneapolis, Minnesota;
4University of Milan, I Division of Infectious Diseases, Luigi Sacco Hospital, Italy;5University of Bari, Clinic of Infectious Diseases, Italy;6Department of Infec-
tious Diseases, Amedeo di Savoia Hospital, Turin, Italy;7Department of Infectious Diseases, Ferrara, Azienda Ospedaliera Universitaria, Italy;8S.M. Goretti
Hospital, La Sapienza University, Polo Pontino, Latina, Italy;9Division of Infectious Diseases, San Martino University Hospital, Genoa, Italy;10University of
Milan, Clinic of Infectious and Tropical Diseases, Department of Medicine, Surgery and Dentistry, San Paolo Hospital, Italy
Preliminary result of this study was presented as oral poster discussion at the XVIII International AIDS Conference, July 18, 2010, Vienna Austria; as poster
presentation at the 52nd Annual Meeting of the American Society of Hematology (ASH), Orlando FL, December 4, 2010; as oral themed discussion at the
18th Conference on Retrovirus and Opportunistic Infections (CROI 2011), Boston, MA, February 27, 2011.
Conflict of Interest: No member of the Writing Group for this report has any financial or personal relationships with people or organizations that could inap-
propriately influence this work, although most members of the group have, at some stage in the past, received funding from a variety of pharmaceutical com-
panies for research, travel grants, speaking engagements, or consultancy fees.
*Correspondence to: Michele Bibas, Clinical Department, National Institute for Infectious Diseases, ‘‘Lazzaro Spallanzani,’’ IRCCS, Via Portuense 292, 00149
Rome, Italy. E-mail: email@example.com
Received for publication 17 February 2012; Revised 26 March 2012; Accepted 13 April 2012
Am. J. Hematol. 87:749–753, 2012.
Published online 26 April 2012 in Wiley Online Library (wileyonlinelibrary.com).
C 2012 Wiley Periodicals, Inc.
American Journal of Hematology
detailed description, and for the relative incidence of malignancies in
this cohort, see d’Arminio Monforte et al. [17,18]. Briefly, patients are
enrolled in the cohort when antiretroviral therapy-naive and followed
over time. By study protocol, blood samples of participants are col-
lected at 4–12-month intervals and shipped to a central repository for
storage. Institutional review boards at each of the sites participating in
study approved the collection of plasma samples, and written informed
consent was obtained from all participants.
For this analysis, information on incident NHL and HL was obtained
prospectively using medical record review. Patients were selected as
cases if ?18 years old, HIV infected at the time of NHL and HL diagno-
sis, and had available serum or plasma samples obtained before the
lymphoma diagnosis and no other major comorbidities [concomitant
malignancies, monoclonal gammopathy of undetermined significance
(MGUS), renal impairment, autoimmune conditions like active rheu-
matic disease, and other chronic inflammatory conditions].
A total of 46 subjects with lymphoma were matched 1:1 with 46
lymphoma-free control patients so that every sample stored before a
lymphoma diagnosis was matched 1:1 with a lymphoma-free control
sample. For each case, the control is chosen randomly among those
members of the cohort who were at risk for developing lymphoma
matching for the duration of time from the date of case sample to the
date of lymphoma diagnosis, according to the statistical procedure for
the design of a case-control study nested in a cohort study. Matching
variables were sex (male and female) and age (stratified as 18–38 and
39–58 years, there were no cases over 58 years old). Besides these
two variables, controls were further matched to cases according to the
duration of time from sample date to the date of diagnosis for cases or
the time of selection for controls, stratified as <2, 2–5, and >5 years.
We adopted a prospective-specimen-collection, retrospective-blinded-
evaluation (PRoBE) design of the study to ensure a high-quality bio-
marker validation [19,20].
Serum FLC analysis. For all patients, serum or plasma stored at
2808C were thawed and analyzed at the National Institute for Infec-
tious Disease ‘‘Lazzaro Spallanzani’’ laboratory. The free light chain
(FLC) assay measures the concentration in the serum of immunoglobu-
lin kappa (j) and lambda (k) light chains that are not attached to a
heavy chain. The sFLC concentrations were determined using the
quantitative FLC assay (FREELITE; The Binding Site, Birmingham, UK)
and performed on a Delta, Radim Nephelometer using kits provided
courtesy of The Binding Site, Italy. The assay separately measure j
sFLC (normal range, 0.33–1.94 mg/dl) and k sFLC (normal range,
0.57–2.63 mg/dl), and j:k FLC ratio (normal range 0.26–1.65) [21,22].
Because the polyclonal nature of free light chains (FLCs) was the focus
of our study, we introduced the j 1 k sum (normal range, 0.59–4.57
mg/dl) as the measurement of choice and as the primary variable stud-
ied. Of interest, all sFLC tests were performed in a single working ses-
sion to avoid as much as possible interlaboratory differences in the
methodological approach and instrumental sensibility. Furthermore, lab-
oratory staff handling and testing specimens were kept blind to the
case-control status of the samples.
Statistical analysis.The odds of lymphoma were estimated using
conditional logistic regression [23,24]. We investigated the univariable
associations between the following variables and the outcome: mode of
HIV transmission, nadir of CD4 cell count, CD4 cell count at sample
time, log10HIV-RNA at sample time, time spent with an undetectable
viral load before the date of sample, HAART exposure at sample time,
and diagnosis/selection time.
In multivariable analysis, we included only variables significant at the
univariable analysis with a P < 0.05. j, k, and j 1 k sFLC values were
used as continuous values in the row scale, as natural log-transformed,
as categorized variables using the quartile distribution and using the
multiples of upper normal limit (UNL) as cutoffs. Nadir CD4, which was
collinear with the CD4 at time of sample, and use of HAART which was
collinear with time spent with a suppressed viral load were not retained
in the final model.
Median sFLC values in sample-cases with a diagnosis of HL were
compared with those in sample-cases with a diagnosis of NHL. Com-
parisons of the median j and k sFLC values between case and con-
trols and between HL and NHL were made by Wilcoxon Mann–Whitney
test. Differences in means (in the natural log scale) were investigated
using linear regression, and the hypothesis that the magnitude of these
differences was varying according to the time distance from the stored
sample was tested including interaction terms in the model. A similar
analysis was conducted using the conditional logistic regression model
in which the association between sFLC and the risk of malignancies
was also assessed separately for NHL and HL. Statistical analysis was
performed using STATA statistical software, release 10, 1999, Stata
Samples selection and sFLC concentration
Of 6,513 ICONA participants enrolled from 1997 to 2007,
86 patients with incident lymphoma were identified. Of
these, 40 were subsequently excluded: 18 for the presence
of confounding comorbidities [(MGUS (two) or other malig-
nancies (four Kaposi), renal dysfunction (eight), autoim-
mune conditions as active rheumatic disease or other
chronic inflammatory conditions (four)]. The reason for
excluding the additional 22 was due to the fact that none of
the identified matched controls had a stored sample at the
required date. Samples are stored routinely, but it is possi-
ble that all vials have been used at specific dates (because
selected for other projects) or have not been stored at that
particular date. It is reasonable to assume that samples are
missing at random and therefore unlikely to have intro-
duced bias in this analysis.
As a result, 46 of the initial 86 patients with incident lym-
phoma were included in this analysis (89.1% men, median
age 37 years). Of them, 16 had HL and 30 patients had
NHL comprising 23 DLBCL, five Burkitt lymphoma, two pri-
mary central nervous system (CNS) lymphoma. The major-
ity had advanced (stage III–IV) lymphoma and more than
half patients were receiving therapy for HIV at the time of
the sample (Table I).
Of interest samples from both patients with HIV-related
lymphoma and lymphoma-free controls had a median value
for j and k higher than the normal upper limit value of the
test , In case-samples the median interquartile range
(IQR) of j 1 k FLCs value were significantly higher: 9.21
(7.33–13.02) than those observed in lymphoma-free control
: 6.11 (4.08–7.54) (P < 0.001). The sFLC assay was used
also to reveal the presence of clonal j- or k- restricted tu-
mor cell population by detecting abnormalities in the physi-
ologic ratio of sFLC. On this respect, the median j:k ratio
was found normal in both patients with HIV-related lym-
phoma and lymphoma-free controls, (only one case sample
had abnormal k:k ratio), confirming the polyclonal nature of
the B-cell population synthesizing the excess of sFLC and
the absence of monoclonal gammopathy in all, but one,
Temporal distribution of samples and sFLC value
When looking at the temporal distribution of samples, we
found 2 years as the median value of the distribution of
time from samples, so we divided samples in early and late
according to time less/equal or more than 2 years. FLCs
concentration were found greatly and significantly increased
TABLE I. Clinical Characteristics of Cases and Controls
(N 5 46)
(N 5 46)P
Age, years, median (IQR)
HCV1, n (%)
Calendar year of
sample, median (IQR)
Years of HIV infection at
date of sample, median (IQR)
CD4 count at sample,
cell/mmc, median (IQR)
HIV-RNA at sample,
log10cp/ml, median (IQR)
Use of ARV at sample, n (%)
Use of HAART at sample, n (%)
Months of viral suppression
before sample, mean (SD)
7.8 (1.4–13.4)9.6 (5.9–14.4)0.017
328 (161–464)576 (362–740) 0.0002
3.70 (2.30–4.69) 1.70 (1.70–3.49)0.0001
Abbreviations: IQR, interquartile range; SD, standard deviation; ARV, antiretrovi-
ral therapy; HAART, highly active antiretroviral therapy.
750 American Journal of Hematology
in cases versus controls when restricting to samples col-
lected ?2 years before the date of diagnosis (the matching
date for controls); median (IQR) j 1 k sFLC in cases 5
10.70 (7.33–13.02); j 1 k sFLC in controls 5 4.85 (4.02–
7.12) P > 0.0001. In contrast, we found smaller differences
between cases and controls when using samples collected
in the >2 years time period. Median (IQR) j 1 k sFLC in
cases 5 6.99 (5.50–9.67); j 1 k sFLC in controls 5 6.74
(6.37–8.52) P > 0.954 (Figure 1).
Risk stratification by type of lymphoma
A similar analysis was conducted using the binary out-
come and conditional logistic regression and after further
stratification by the type of lymphoma (NHL and HL) When
considering the elevation of j 1 k sFLC more than 1.5-fold
the UNL in the 0–2 years period, we found for NHL an OR
of 15 (95% CI 1.98–113.56) compared with the OR of 3.0
(95% CI 0.98–11.08) for HL. A different pattern resulted in
the >2 years stratum: for j 1 k sFLC > fold the 1.5 UNL,
we found an OR of 3.0 for NHL and an OR not calculable
(due to numerical limitations) for HL.
Predictors of lymphoma
Overall, at univariable logistic regression analysis, com-
pared to people with a value below or equal the ULN, sFLC
were significantly associated with the risk of lymphoma.
Other important predictors associated with a higher risk of
lymphoma were lower nadir of CD4, lower count of CD4 at
sample and higher log10HIV-RNA at sample. The use of
HAART at sample and longer time spent in months with
undetectable HIV-RNA before the sample date were predic-
tive of lymphoma risk (Table II). Multivariable analysis con-
firmed the association between higher risk of lymphoma in
patients and sFLC values above the UNL. Furthermore, in
all multivariable models, we found that for j 1 k sFLC the
reduction of lymphoma risk was associated with periods of
undetectable HIV viremia lasting more than 6 months. In
contrast, no significant association with CD4 cell count at
sample time was found (Table III).
sFLC in controls
A key question addressed in this study was to observe
associations with sFLC among the controls particularly with
respect to: (1) use of HAART at sample; (2) months of viral
suppression at sample; (3) HIV viral load at sample; (4)
CD4 cell count at sample. This cross-sectional analysis
showed that: (1) Untreated patients have a significantly
higher j 1 k sFLC value compared to patients who were in
treatment with HAART independently of being cases or
controls (Table IVa). (2) A short time of undetectable HIV
viremia was associated with a statistically significant higher
j 1 k sFLC (Table IVb). (3) The detection of high viral load
at sample was associated with a statistically significant
higher j 1 k sFLC (Table IVc). (4) Controls and cases who
had lower CD4 cell count had a significantly higher j 1 k
sFLC (Table IVd).
Several groups have examined the association between
a variety of biomarkers of functional perturbations of B-cell
(notably CD30, serum globulin level, the presence of a
monoclonal protein) and the risk of developing lymphoma
[8,11,25]. Landgren et al.  first reported the association
between elevation of sFLC and the risk of HIV-related NHL
in HIV-infected individuals. Our results confirm these find-
ings; however, some important differences should be high-
lighted. In particular, unlike in the analysis by Landgren, we
with conflicting results
TABLE II. Crude OR of Lymphoma Estimated by Logistic Conditional
Regression of All Covariates Analyzed at Univariable Level
Crude OR (95% CI)P
k 1 k–sFLC
Mode of HIV transmission
Nadir of CD4 ?200 (cell/mm3)
CD4 cell count at sample
(per 100 cells higher) (cell/mm3)
On HAART at sample time
Time of HIV-RNA undetectable (months)
Log10HIV-RNA, mean (SD)
0.22 (0.07–0.66) 0.007
Abbreviations: UNL, upper normal limit; HAART, highly active antiretroviral ther-
apy; sFLC, serum free light chain; OR, odd ratio; CI, confidence interval; SD,
UNL for j: 1.94 mg/dL; k: 2,63 mg/dL; j 1 k: 4.57 mg/dl.
TABLE III. Multivariable Conditional Logistic Regression Models Including
j 1k–sFLC as Categorical Variables Using the Upper Normal Limit as
OR (95% CI)P
j 1 k-sFLC
Time of HIV-RNA undetectable (months)
CD4 cell count at sample
(per 100 cells higher), cell/mm3
HIV-RNA (cp/ml) (per 1 log10increase)
Abbreviations: UNL, upper normal limit; HAART, highly active antiretroviral ther-
apy; j 1 k-sFLC, total serum free light chains; OR, odd ratio; CI, confidence inter-
val; j 1 k: 4.57 mg/dl.
time from sample (ST) less/equal or more than 2 years(YRS). Black points repre-
sents median value.
Kappa 1 lambda values distribution in cases and controls according to
American Journal of Hematology751
excluded by inclusion criteria of all patients in who elevated
serum FLCs could be due to the presence of other comor-
bidities such as progressive renal impairment, MGUS, con-
comitant malignancies, autoimmune diseases, and other
chronic inflammatory conditions [26,27]. Moreover, partici-
pants in the Landgren analysis were mainly non-HAART
treated whereas near a half of our lymphoma patients and
more than three quarters of controls were receiving HAART
at the time of the sample so that we were able to evaluate
the association with time with a suppressed viral load and
with CD4 cell count, achieved on HAART on this biomarker.
Samples from both patients with HIV-related lymphoma
and lymphoma-free controls had an overall j 1 k sFLC
higher than UNL. This is probably due to the HIV-specific
activation of B cells induced by HIV infection and may indi-
cate the level of perturbations/disruption of B-cell functions.
Further analysis correlating levels of FLC with parameters
measuring B-cell functions are warranted to support this hy-
pothesis. sFLC might have several applications in the man-
agement of HIV infection which remain unresolved [28,29].
For example, it could be combined with the classic CD4 T-
Cell count in evaluating overall immune recovery during
HAART or to provide information to determine which HIV-
infected individual might benefit most from early initiation of
HAART because at high risk of malignancies.
Notably, our results emphasize that when restricting the
analysis to samples stored over the previous 2 years, there
was the biggest difference in sFLC values between patients
with or without lymphoma. If, as we hypothesize, polyclonal
FLCs are a marker of B-cell disruption, this result suggest
that it may take up to 2 years until this disruption becomes
of etiological significance. However, this observation needs
to be confirmed in other studies entailing the collection of
sequential blood samples as well as biomarkers measuring
the functional perturbations of B-cell.
Interestingly, we documented that, after adjusting for
sFLC value, the longer the cumulative time spent with HIV
viremia under the threshold of detectability, the lower was
the risk of lymphoma. As the effect was independent of
CD4 count (both at nadir or at sample), the protective effect
of longer duration of undetectable viremia on the risk of
lymphoma may be explained by an effective concomitant
decrement of other disturbances of immune function, such
as chronic immune stimulation with a consecutive reduction
of the ongoing polyclonal B-cell activation/dysfunction.
Importantly, we did not found a significant independent
association between CD4 cell count and the risk of lymphoma.
In our model, CD4 cell count at sampling and nadir of CD4
was associated with risk of lymphoma only in univariable anal-
ysis. In multivariate model, the strong effect of duration of sup-
pressive HIV viremia contributed to dilute the global effect of
CD4. Moreover, CD4 count were measured some time before
the date of diagnosis of lymphoma, and this may further
explain the low predictive value of this marker.
In addition, the inclusion of subjects with HL in this study
gave us the opportunity to speculate about the different
role of SFLC according to the lymphoma type. In fact, the
elevated sFLC seems to be more associated with the risk
of NHL than HL. These differences might be due to the dif-
ferent pathogenetic mechanisms, but, because of the small
sample size of the HL group, we cannot exclude that these
findings were due to chance or to lack of statistical power.
NHL is notably more related to the prolonged immuno-
suppression [8,30]. B-cell disruption might contribute to the
NHL pathogenesis through the reactivation of a previous
infection with Epstein–Barr virus (EBV) resulting in the
long-term stimulation and the proliferation of impaired B
lymphocytes, ultimately leading to the development of HIV-
related NHL . We hypothesize that the marked and pro-
longed elevations of sFLC in HIV-related NHL, despite an
effective HAART and at a relatively preserved CD4 cell
count, may reflect the persistence of a profound B-cell dys-
function not completely restored by the antiretroviral ther-
apy. Those patients may be at higher risk of developing
NHL and should be monitored more closely over the time.
In contrast, some degree of immune competence is
required for HL to develop. Biggar et al.  reported a 14-
fold increase in incidence of HL in patients with a CD4 cell
count of 150–199 cell/ml compared with those with <50
CD4 cells/ml. Those findings have been confirmed also by
larger collaborative groups [6,33,34]. In our analysis, we
tested the interaction between sFLC and CD4 count, but
the effect of sFLC seems to be unchanged by CD4 level.
Based on this evidence, the lower elevation of sFLC in
HIV-related HL compared with NHL patients might indicate
a less generalized perturbation of the B-cell compartment
probably influencing more specifically the tumor microenvir-
onment, and mechanisms different from immune B-cell
stimulation may have a role in the development of HL.
This work has several limitations that need to be consid-
ered. First of all, the study might be underpowered espe-
cially for examining NHL and HL separately. Despite this,
our results gave us suggestions that there is a difference
between them. However, this observation needs to be con-
firmed in other studies. We did not have information regard-
ing tumor EBV status which could be an important unmeas-
ured confounder. Another limitation, common to other stud-
ies with similar case-control retrospective design, is the
lack of central blinded pathology review of lymphoma
cases. However, some strengths of this study should be
mentioned. First of all, the ICONA study is a well-character-
ized cohort with a central repository of plasma samples in
which malignancies are collected in a standardized manner.
Moreover, the PRoBE design the assessment of the assay
in a single experienced clinical laboratory and in a single
session guarantee a reduced risk of measurement error.
In conclusion, we showed that elevated sFLC is a strong
and sensitive marker of the risk of developing lymphoma (par-
ticularly for NHL) in HIV1individuals for a given gender, age,
CD4 count, and duration of viral suppression on HAART.
Because sFLC are easily measured, they are potentially can-
didate biomarkers for screening and prognosis in HIV-related
lymphomas, and, therefore, it is important that our results are
confirmed in other settings. Further studies, also evaluating
cytokine activation, are needed to better understand the path-
ogenetic mechanism that underlies the B-cell activation
reflected by sFLC elevation before lymphoma.
TABLE IV. sFLC in Controls
(a) HAART at sample
j 1 k-sFLC, median (IQR)
(b) Months of viral suppression
4.86 (3.65–6.76) 9.27 (6.72–13.35)0.001
j 1 k-sFLC, median (IQR)
(c) HIV-RNA (copies/ml)
7.70 (6.64–12.52) 4.85 (3.65–7.10) 0.005
j 1 k-sFLC, median (IQR)
(d) CD4 (cell/mm3)
4.86 (3.65–6.76)7.54 (4.59–12.13)0.013
j 1 k-sFLC, median (IQR) 7.10 (6.00–12.13) 4.58 (3.75–6.74)0.017
Median (IQR) Values of j 1 k in Different Categories of Risk factors: (a) HAART
at sample, (b) months of viral suppression, (c) HIV-RNA copies at sample, (d)
CD4 count at sample. Comparisons between two categories were made by Wil-
coxon Mann-Whitney test.
Abbreviations: HAART, highly active antiretroviral therapy; j 1k-sFLC, total se-
rum free light chain.
UNL j 1 k: 4.57 mg/dl.
752 American Journal of Hematology
Acknowledgment Download full-text
The authors thank Barbara Amoroso, MD, PhD, for her
Author Contributions: Michele Bibas, Maria Paola Trotta
and Andrea Antinori was responsible for study design and
interpretation of findings; drafted and revised the manu-
script critically for important intellectual content and submis-
sion. Alessandro Cozzi-Lepri revised the manuscript crit-
ically for important intellectual content and coordinated the
statistical analysis. Patrizia Lorenzini set up the database
and performed the statistical analyses. Cristina Di Giacomo
performed the laboratory analysis. Giuliano Rizzardini,
Gioacchino Angarano, Carmela Pinnetti, Pietro Caramello,
Laura Sighinolfi, Claudio Maria Mastroianni, Giovanni Maz-
zarello, AntoninoDi Caro,
d’Arminio Monforte and Andrea Antinori coordinated the
study. All authors read and approved the final manuscript.
ICONA Foundation Study Group:
Governing body M. Moroni (Chair), G. Carosi, R.
Cauda, F. Chiodo, A. d’Arminio Monforte, G. Di Perri, M.
Galli, R. Iardino, G. Ippolito, A. Lazzarin, R. Panebianco, G.
Pastore and C. F. Perno.
Steering committee A. Ammassari, A. Antinori, C. Arici,
C. Balotta, P. Bonfanti, M. R. Capobianchi, A. Castagna, F.
Ceccherini-Silberstein, A. Cozzi-Lepri, A. d’Arminio Mon-
forte, A. De Luca, C. Gervasoni, E. Girardi, S. Lo Caputo,
R. Murri, C. Mussini, M. Puoti, and C. Torti.
Participating physicians and centres M. Montroni, A.
Costantini, F. Frontini (Ancona); A. Giacometti, A. Riva
(Ancona); G. Angarano, C. Carrisa (Bari); F. Maggiolo, G.
Lazzari (Bergamo); P. Viale, G. Verucchi, E. Vanini (Bolo-
gna); G. Carosi, C. Torti, C. Minardi (Brescia); T. Quirino, C
Abeli (Busto Arsizio); P. E. Manconi, P. Piano (Cagliari); J.
Vecchiet, K. Falasca (Chieti); L. Sighinolfi, D. Segala (Fer-
rara); F. Mazzotta, S. Lo Caputo (Firenze); G. Cassola, R.
Piscopo (Genova); C. Viscoli, G. Mazzarello, A. Alessan-
drini (Genova); C. Mastroianni, V. Belvisi (Latina); P. Bon-
fanti, I. Caremma (Lecco); A. Chiodera, P. Castelli (Mac-
erata); M. Galli, A. Ridolfo, R. Piolini (Milano); G. Rizzar-
dini, P. Zucchi (Milano); A. d’Arminio Monforte, T. Formenti
(Milano); A. Lazzarin, A. Castagna, S. Salpietro (Milano);
M. Puoti, C. Moioli (Milano); R. Esposito, C. Mussini, L.
Bisio (Modena); A. Gori, G. Lapadula (Monza); A. Chirianni
(Napoli); N. Abrescia, M. De Marco (Napoli); F. Baldelli, B.
Belfiori (Perugia); G. Parruti, F Sozio (Pescara); G. Mag-
nani, M. Ursitti (Reggio Emilia); R. Cauda, A. De Luca, A.
Marzocchetti (Roma); P. Narciso, L. Gallo (Roma); A. Anti-
nori, R. Acinapura (Roma); V. Vullo, M. Lichtner, G. Tebano
(Roma); M. Andreoni, M. Capozzi (Roma); V. Tozzi, R. Lib-
ertone, S. Cerilli (Roma); A. M. Cattelan (Rovigo); M. S.
Mura, G. Madeddu (Sassari); G. Di Perri, S. Bonora, M.
Sciandra (Torino); P. Caramello, M. Guastavigna (Torino);
G. Pellizzer, V. Manfrin (Vicenza).
1. Grulich AE, van Leeuwen MT, Falster MO, Vajdic CM. Incidence of cancers in
people with HIV/AIDS compared with immunosuppressed transplant recipi-
ents: A meta-analysis: Lancet 2007;370:59–67.
2. Simard EP, Engels EA. Cancer as a cause of death among people with AIDS
in the United States. Clin Infect Dis 2010;51:957–962.
3. International Collaboration on HIV and Cancer. Highly active antiretroviral
therapy and incidence of cancer in human immnunodeficiency virus infected
adults. J Natl Cancer Inst 2000;92:1823–1830.
4. Simard EP, Pfeiffer RM, Engels EA. Cumulative incidence of cancer among
individuals with acquired immunodeficiency syndrome in the United States.
5. Bohlius J, Schmidlin K, Boue ´ F, et al. HIV-1-related Hodgkin lymphoma in the
era of combination antiretroviral therapy: Incidence and evolution of CD41 T-
cell lymphocytes. Blood 2011;117:6100–6108.
6. Clifford GM, Rickenbach M, Lise M, et al. Hodgkin lymphoma in the Swiss
HIV Cohort Study. Blood 2009;113:5737–5742.
7. Stebbing J, Gazzard B, Mandalia S, et al. Antiretroviral treatment regimens
and immune parameters in the prevention of systemic AIDS-related non-
Hodgkin’s lymphoma. J Clin Oncol 2004;22:2177–2183.
8. Engels EA, Pfeiffer RM, Landgren O, Moore RD. Immunologic and virologic
predictors of AIDS-related non-Hodgkin lymphoma in the highly active antire-
troviral therapy era. J Acquir Immune Defic Syndr 2010;54:78–84.
9. Cagigi A, Nilsson A, De Milito A, Chiodi F. B cell immunopathology during
HIV-1 infection: Lesson to learn for HIV-1 vaccine design. Vaccine 2008;26:
10. Moir S, Buckner CM, Ho J, et al. B cells in early and chronic HIV infection:
Evidence for preservation of immune function associated with early initiation
of antiretroviral therapy. Blood 2010;116:5571–5579.
11. Vendrame E, Martı ´nez-Maza O. Assessment of pre-diagnosis biomarkers of
immune activation and inflammation: Insights on the etiology of lymphoma. J.
Proteome Res 2011;10:113–119.
12. Breen EC, Fatahi S, Epeldegui M, Boscardin WJ, Detels R, Martı ´nez-Maza
O. Elevated serum soluble CD30 precedes the development of AIDS-associ-
ated non-Hodgkin’s B cell lymphoma. Tumour Biol 2006;27:187–194.
13. Landgren O, Goedert JJ, Rabkin CS, et al. Circulating serum free light chains
as predictive markers of AIDS-related lymphoma. J Clin Oncol 2010;28:
14. Maurer MJ, Micaleff IN, Cerhan JA, et al. Elevated serum free light chains
are associated with event-free and overall survival in two indipendent cohorts
of patients with diffuse large B-cell lymphoma. J Clin Oncol 2011;29:
15. Thompson CA, Maurer MJ, Cerhan JR, et al. Elevated serum free light chains
are associated with inferior event free and overall survival in Hodgkin lym-
phoma. Am J Hematol 2011;86:998–1000.
16. Morabito F, De Filippi R, Laurenti L, et al. The cumulative amount of serum
free light chain is a strong prognosticator in chronic lymphocytic leukemia.
17. d’Arminio Monforte A, Lepri AC, Rezza G, et al. Insights into the reasons for
discontinuation of the first highly active antiretroviral therapy (HAART) regi-
men in a cohort of antiretroviral naı ¨ve patients. AIDS 2000;14:499–507.
18. Prosperi MC, Cozzi-Lepri A, Castagna A, et al. Incidence of malignancies in
HIV-infected patients and prognostic role of current CD4 cell count: Evidence
from a large Italian cohort study. Clin Infect Dis 2010;50:1316–1321.
19. Pepe MS, Feng Z, Janes H, et al. Pivotal evaluation of the accuracy of a bio-
marker used for classification of prediction: Standards for study design. J Natl
Cancer Inst 2008;100:1432–1438.
20. Hanash SM, Baik CS, Kallioniemi O. Emerging molecucal biomarkers-blood
based strategies to detect and monitor cancer. Nat Rev Clin Oncol 2011;
21. Katzmann JA, Clarck RJ, Abraham RS, et al. Serum reference intervals and
diagnostic ranges for free kappa and free lambda immunoglobulin light
chains: Relative sensitivity for detection of monoclonal light chains. Clin
22. Bradwell AR.Immunoglobuliln free light chains and their analysis: Normal
ranges and reference interval. In:
Hevylite), 6th ed. University of Birmingham: The Binding Site Group; 2010;
23. Hosmer DW, Lemeshow S. Applied Logistic Regression, 2nd ed.; John Wiley
& Sons, Inc, 2000.
24. Liang KY, Zeger SL. Longitudinal data analysis using generalized linear mod-
els. Biometrika 1986;73:13–22.
25. Grulich AE, Wan X, Law MG, et al. B-cell stimulation and prolonged immune
deficiency are risk factors for non-Hodgkin’s lymphoma in people with AIDS.
26. Gottenberg JE, Aucouturier F, Goetz J, et al:
light chain assessment in rheumatoid arthritis and primary Sjogren’s syn-
drome. Ann Rheum Dis 2007;66:23–27.
27. Diamantidis MD, Ioannidou-papagiannaki E, Ntaios G: Novel extended refer-
ence range for serum kappa/lamda free light chain ratio in diagnosing mono-
clonal gammopathies in renal insufficient patients. Clin Biochem 2009;42:
28. Hammer SM, Eron JJ Jr, Reiss P, et al. Antiretroviral treatment of adult HIV
infection: 2008 recommendations of the International AIDS Society-USA
panel. JAMA 2008;300:555–570.
29. Kitahata MM, Gange SJ, Abraham AG, et al. Effect of early versus
deferred antiretroviral therapy for HIV on survival. N Engl J Med 2009;360:
30. Bower M, Fisher M, Hill T, et al. CD4 counts and the risk of systemic non-
Hodgkin’s lymphoma in individuals with HIV in the UK. Haematologica 2009;
31. Bibas M, Antinori A. EBV and HIV-related lymphoma. Mediterr J Hematol
Infect Dis 2009;1:e2009032.
32. Biggar RJ, Jaffe ES, Goedert JJ, et al. Hodgkin lymphoma and immunodefi-
ciency in person with HIV/AIDS. Blood 2006;108:3786–3791.
33. Bohlius J, Schmidlin K, Boue ´ F, et al. HIV-1-related Hodgkin lymphoma in the
era of combination antiretroviral therapy: Incidence and evolution of CD41 T-
cell lymphocytes. Blood 2011;117:6100–6108.
34. Ayoub JP, Palmer JL, Huh Y, et al. Therapeutic and prognostic implications of
peripheral blood lymphopenia in patients with Hodgkin’s disease. Leuk lym-
Serum Free Light Chain Analysis (plus
Serum immunoglobulin free
American Journal of Hematology753