Anaplastic Glioma: How to Prognosticate Outcome and Choose a Treatment Strategy

VOLUME 27 䡠 NUMBER 35 䡠 DECEMBER 10 2009
Anaplastic Glioma: How to Prognosticate Outcome
and Choose a Treatment Strategy
Lisa M. DeAngelis, Department of Neurology, Memorial-Sloan Kettering Cancer Center, New York, NY
See accompanying articles on pages 5874 and 5881
Anaplastic gliomas are classified by the WHO as grade 3 malignant tumors and include the anaplastic astrocytoma, anaplastic oligodendroglioma, and anaplastic oligoastrocytoma or mixed glioma.
These highly aggressive tumors often occur in young adults and typically recur or progress to a grade 4 glioblastoma within several years of
diagnosis, despite treatment with surgery, radiotherapy, and chemotherapy. There is some evidence that anaplastic glioma is a molecular
precursor to glioblastoma.1 However, these tumors are uncommon
(anaplastic astrocytoma accounts for only 3.2% and anaplastic oligodendroglioma 1.2% of primary brain tumors, compared with the
20.3% incidence of glioblastoma2). Furthermore, they are often heterogenous, harboring a more malignant focus not sampled by the
neurosurgeon, as these tumors are often diffuse and not amenable to
gross total resection. For all these reasons, the anaplastic gliomas were
included with glioblastomas in earlier studies examining new therapeutics and their numbers were usually too small to allow for valid
subgroup analysis.3 About 20 years ago, differences in tumor biology
suggested that the anaplastic glioma may be sufficiently different from
glioblastoma to warrant independent investigation of novel therapies.
Two articles in this issue of Journal of Clinical Oncology highlight some
of the unique features of anaplastic gliomas and how they may differ
from the glioblastoma.4,5
Anaplastic oligodendroglioma was the first to be recognized as a
discrete subgroup uniquely sensitive to chemotherapy and clearly
different from the anaplastic astrocytoma; its chemosensitivity appears linked to loss of heterozygosity for chromosomes 1p and 19q.6
While the anaplastic oligodendroglioma is distinct from the anaplastic
astrocytoma, the anaplastic mixed glioma has been variably reported
to have either a prognosis intermediate between the two subtypes or to
be more closely linked with one or the other. Codeletion of 1p and 19q
may be more predictive of behavior than histology, and two large
international randomized trials are about to open based on classification of anaplastic gliomas by their 1p/19q status and not by their
pathologic appearance. Given the chemosensitivity of anaplastic oligodendroglioma, it was a surprise when two large randomized control
trials comparing radiation therapy (RT) alone to RT plus adjuvant
chemotherapy with procarbazine, lomustine, and vincristine (PCV)
or neoadjuvant PCV failed to show that chemotherapy improved
survival.6,7 Chemotherapy did significantly prolong disease-free survival, and most patients randomly assigned to RT alone received
chemotherapy at progression, complicating interpretation of the survival data. In this issue of JCO, van den Bent et al4,7 have returned to
Journal of Clinical Oncology, Vol 27, No 35 (December 10), 2009: pp 5861-5867
their data from the European Organisation for Research and Treatment of Cancer study and re-analyzed tumor specimens from 152 of
the 368 patients enrolled for methylation, and therefore inactivation,
of the MGMT promoter, a potential mechanism of chemosensitivity.
Unexpectedly, their data show that MGMT promoter methylation is
an independent prognostic factor, conferring better outcome even if
initial treatment does not include an alkylating agent. However, the
authors do not address the observation that patients whose tumors
had a methylated promoter and received PCV in addition to RT had
the best progression-free survival and overall survival, suggesting that
MGMT promoter methylation may be both a prognostic and predictive marker. In addition, those whose tumor had an unmethylated
promoter also did better when chemotherapy was incorporated into
initial treatment. In contrast, MGMT promoter methylation was not a
prognostic factor in their 40 patients whose tumors were re-classified
as glioblastoma on central pathologic review. These findings differ
from Hegi et al,8 who described MGMT promoter methylation as a
predictor in glioblastoma of response to temozolomide.
The results of these studies highlight the uncertainty regarding
the optimal treatment of patients with anaplastic oligodendroglioma.
Even among experienced neuro-oncologists, there is a wide range of
opinion regarding the initial treatment, often, but not always, influenced by 1p/19q status.9 Now, MGMT status may be a critical molecular characteristic but does not appear to be a predictor of therapeutic
response or even a determinant of treatment choice.
Treatment of the anaplastic astrocytoma has been less variable.
This tumor is more resistant to therapy and patients have a shorter
median survival of only 2 to 3 years, compared with 5 years for
anaplastic oligodendroglioma. Most physicians in the United States
treat patients with maximal safe resection and involved field radiotherapy with concurrent and adjuvant temozolomide, identical to the
regimen now considered the standard of care for glioblastoma.10
However, the potential benefit of adding chemotherapy in these
patients has never been established, although temozolomide was
initially granted accelerated approval by the US Food and Drug
Administration based on its efficacy in patients with recurrent anaplastic astrocytoma. Therefore, better information is essential to improving treatment for these patients.
In this issue of JCO, a large German multicenter randomized
controlled trial analyzed two different therapeutic approaches
to patients with newly diagnosed anaplastic glioma.5 The investigators employed a highly unusual study design. Patients were
© 2009 by American Society of Clinical Oncology
Downloaded from on September 17, 2014. For personal use only. No other uses without permission.
Copyright © 2009 American Society of Clinical Oncology. All rights reserved.
randomly assigned to receive either radiation or chemotherapy,
and those randomized to receive chemotherapy were randomly
assigned again to either temozolomide or the PCV regimen. When
patients in the radiotherapy arm experienced relapse (defined by
the MacDonald criteria), they were randomly assigned to receive
either PCV or temozolomide. Patients in the primary chemotherapy arm received radiotherapy on relapse, although approximately
20% received a second course of chemotherapy, delaying RT in this
subgroup. Only after both treatment modalities had been delivered
and demonstrated failure was the primary end point (time to
treatment failure) reached. Thus, the primary end point includes
time to second or even third progression in some patients. The
result for all groups was essentially the same, regardless of the
modality employed as initial therapy. Temozolomide and PCV
were equally efficacious, although PCV was more toxic.
Unfortunately, the interpretation of this study is markedly compromised by its design. The authors do not describe their hypothesis
for the study or how such a complicated trial structure would address
their question. In the absence of pre-existing data suggesting the sequence of therapeutic modalities should affect outcome, the study was
destined to show equivalence between the two treatment arms. Only if
treatment sequence affected tumor progression after all therapy was
administered could there be a difference between the study arms.
Therefore, the results of the trial could have been predicted based on
its design, and the data fail to guide future treatment decisions for
these patients.
Close examination of the results reveals that initial treatment
selection may matter. Median progression-free survival, a secondary
end point, was also similar between treatment groups, although the
data suggest that time to progression after RT may be longer than after
chemotherapy; at 54 months follow-up, 77.8% of patients had completed salvage RT, whereas only 48% had completed salvage chemotherapy. Furthermore, initial RT yielded more complete responses,
partial responses, and stable disease than did initial chemotherapy,
suggesting superiority of RT. It is not clear whether failure of chemotherapy was more rapid in the astrocytoma than the oligodendroglioma group.
The study nicely confirms the importance of molecular markers
in patients with anaplastic gliomas. Loss of heterozygosity of 1p/19q
was a predictor of better outcome. Patients with anaplastic oligodendroglioma and anaplastic mixed glioma had an identical time to treatment failure and progression-free survival, but we do not know what
proportion of patients within these two categories was 1p/19q codeleted. This study also confirmed the importance of IDH 1 mutations as
an important prognostic factor, which is newly recognized to confer a
better prognosis.11,12 In this study, like the report by van den Bent
et al,4,7 patients whose tumors had MGMT promoter methylation did
better regardless of initial therapy.
Both groups of authors are to be congratulated on completing large
randomized trials enrolling 318 and 368 patients, respectively, with anaplastic gliomas, confirmed by central pathology review (which led to
substantial reclassification), and acquiring tissue for study of molecular correlates. However, neither study clarifies the best therapeutic approach to these patients using our current conventional
armamentarium, but based on these studies, treatment decisions
should not rest on MGMT promoter methylation. More specific
information awaits completion of the new studies about to begin
using 1p/19q classificiation for the first time.
The author(s) indicated no potential conflicts of interest.
1. Kleihues P, Burger PC, Aldape KD, et al: Glioblastoma, in Louis DN, Ohgaki
H, Wiester OD, et al (eds): WHO Classification of Tumours of the Central Nervous
System (ed 4). Lyon, France, IARC, 2007, pp 33-49
2. Central Brain Tumor Registry of the United States: Statistical Report:
Primary Brain Tumors in the United States, 1998-2002. Hinsdale, IL, Central Brain
Tumor Registry of the United States, 2005
3. Selker RG, Shapiro WR, Burger P, et al: The Brain Tumor Cooperative
Group NIH Trial 87-01: A randomized comparison of surgery, external radiotherapy, and carmustine versus surgery, interstitial radiotherapy boost, external
radiation therapy, and carmustine. Neurosurgery 51:343-355, 2002
4. van den Bent MJ, Dubbink HJ, Sanson M, et al: MGMT promoter
methylation is prognostic but not predictive for outcome to adjuvant PCV
chemotherapy in anaplastic oligodendroglial tumors: A report from EORTC Brain
Tumor Group study 26951. J Clin Oncol 27:5881-5886, 2009
5. Wick W, Hartmann C, Engel C, et al: NOA-04 randomized phase III trial of
sequential radiochemotherapy of anaplastic glioma with procarbazine, lomustine,
and vincristine or temozolomide. J Clin Oncol 27:5874-5880, 2009
6. Cairncross G, Berkey B, Shaw E, et al: Phase III trial of chemotherapy plus
radiotherapy compared with radiotherapy alone for pure and mixed anaplastic
oligodendroglioma: Intergroup Radiation Therapy Oncology Group Trial 9402.
J Clin Oncol 24:2707-2014, 2006
7. van den Bent MJ, Carpentier AF, Brandes AA, et al: Adjuvant procarbazine,
lomustine, and vincristine improves progression-free survival but not overall
survival in newly diagnosed anaplastic oligodendrogliomas and oligoastrocytomas: A randomized European Organisation for Research and Treatment of
Cancer phase III trial. J Clin Oncol 24:2715-2722, 2006
8. Hegi ME, Diserens A-C, Gorlia T, et al: MGMT gene silencing and benefit
from temozolomide in glioblastoma. N Engl J Med 352:997-1003, 2005
9. Abrey LE, Louis DN, Palelogos N, et al: Survey of treatment recommendations for anaplastic oligodendroglioma. Neuro Oncol 9:314-318, 2007
10. Stupp R, Mason WP, van den Bent MJ, et al: Radiotherapy plus concomitant
and adjuvant temozolomide for glioblastoma. N Engl J Med 352:987-996, 2005
11. Zhao S, Lin Y, Xu W, et al: Glioma-derived mutations in IDH1 dominantly
inhibit IDH1 catalytic activity and induce HIF-1alpha. Science 324:261-265, 2009
12. Yan H, Parsons DW, Jin G, et al: IDH1 and IDH2 mutations in gliomas.
N Engl J Med 360:765-773, 2009
DOI: 10.1200/JCO.2009.24.5985; published online ahead of print at on November 9, 2009
■ ■ ■
Contralateral Breast Cancer in BRCA1/BRCA2
Mutation Carriers: The Story of the Other Side
Judy E. Garber and Mehra Golshan, Dana Farber Cancer Institute; Brigham and Women’s Hospital, Boston, MA
See accompanying article on page 5887
© 2009 by American Society of Clinical Oncology
Downloaded from on September 17, 2014. For personal use only. No other uses without permission.
Copyright © 2009 American Society of Clinical Oncology. All rights reserved.
The identification of the BRCA1 and BRCA2 breast/ovarian cancer susceptibility genes in the mid-1990s and the rapid introduction of
genetic testing thereafter have had impact beyond initial expectations.
Thousands of individuals and families have been tested, with notable
involvement in ongoing research by many. Large-scale collaborative
efforts by investigators around the world (eg, Consortium of Investigators of Modifiers of BRCA1/2 and Breast Cancer Association Consortium) strive to refine breast and ovarian cancer risk estimates for
unaffected mutation carriers by examining mutation-specific risk,
lifestyle factors, and growing lists of modifier genes. Among the important goals of this work is the provision of more precise and stable
estimates of specific risks of primary and contralateral breast, ovarian,
and other associated cancers, and the efficacy of risk-reducing interventions. The data are intended to be used to help guide decisions as to
when to supplement or replace surveillance with medical or surgical
risk reduction and their attendant reconstruction, menopause management, and psychological and emotional adjustments.
In the early days of BRCA1/2, a woman with a breast cancer
diagnosis might have been considered most important as the index
case through whom family members could be identified for genetic
risk determination and management. Now, however, it can be argued
that knowledge of BRCA1/2 mutation status is beginning to influence
cancer management for patients with breast and ovarian cancers.
Early phase trial data presented at the 2009 American Society of Clinical Oncology meeting demonstrating sensitivity of metastatic breast
and ovarian cancers in women with germline BRCA1/2 mutations to a
novel PARP inhibitor, and of newly diagnosed breast cancers in
BRCA1 mutation carriers to the DNA-damaging agent cisplatin,
raise the possibility of oncologic therapies specifically targeted to
this group.1-3 The fact that BRCA1/2 mutations status was a required eligibility criterion for enrollment into these trials, which
were conducted in multiple different countries, is significant. Consideration of upcoming clinical trial participation with promising
targeted agents may drive more genetic testing among women with
metastatic breast and ovarian cancers. If it is ultimately shown that
BRCA1/2-associated breast and ovarian cancers should be treated
differently from their sporadic counterparts, then genetic testing
would become part of the standard assessment of women at diagnosis of metastatic disease, and ultimately at initial breast or ovarian cancer diagnosis.
Has that time already come? The possibility that primary therapy
of newly diagnosed breast cancers could be influenced by knowledge
of BRCA1/2 mutation status has been considered by surgeons and
radiation oncologists. Currently, although ideal data from long-term
follow-up of prospective cohorts are lacking, there appears to be no
support for early concerns either that BRCA1/2-mutation carriers are
significantly more susceptible to the carcinogenic effects of radiation
in their normal cells, or that their breast cancers are more resistant to
therapeutic effects of ionizing radiation.4,5 There have been, however,
a number of series consistently demonstrating an increased risk of
second primary breast cancers in BRCA1/2-mutation carriers.4-7 Several groups have documented the incorporation of such data into
decisions about surgical therapy, in which mutation status influenced
women’s choices of bilateral mastectomy over lumpectomy plus radiation therapy at breast cancer diagnosis, and have not shown excessive distress.8-11
Should bilateral mastectomy at the time of a breast cancer diagnosis be recommended to all women with BRCA1/2 mutations? The
contribution of Graeser et al12 in this issue of the Journal of Clinical
Oncology provides important data to inform this issue. The investigators utilize the German Consortium for Hereditary Breast and Ovarian Cancer registry data, prospectively collected over more than a
decade, to provide estimates that more precisely quantify the risk of
contralateral invasive breast cancer among women with a BRCA1 or
BRCA2 germline mutation based on age at first breast cancer diagnosis. Of concern, they found that overall contralateral risk at 25 years
after first breast cancer diagnosis reached nearly 50% for both BRCA1
and BRCA2 mutation carriers, and continued to increase thereafter.
As in other cohorts, women with BRCA1 mutations were significantly
younger at both first and second breast cancer diagnoses, which is
important since younger age at first breast cancer diagnosis was associated with higher contralateral risk. In the youngest group, women
first diagnosed before age 40 years, contralateral risk after 25 years was
62.9% (95% CI, 50.4 to 75.4). However, for BRCA1 carriers whose first
breast cancer occurred after age 50, the risk of second primary was
19.6%, and for their BRCA2 counterparts, 16.7%. General population
estimates of contralateral breast cancer risk are in this range.13 The
authors provide the data in a particularly useful table with estimates by
mutated gene, age at diagnosis, and intervals from first breast cancer diagnosis.
These data serve at least two important purposes. First, for
women with BRCA1/2 mutations diagnosed at young ages, they provide powerful figures that should compel the breast cancer care team
to consider the issue of management of the opposite breast. Since most
women with a new breast cancer will not know their BRCA1/2 status at
diagnosis, surgeons in particular must recognize that a patient could
be a mutation carrier, based on age at diagnosis, family history, ethnicity, and possibly histologic features, and offer to refer for genetic
testing as appropriate. Of course, some women will not be ready to
consider genetic testing or prophylactic mastectomy at diagnosis,
but should be offered the opportunity to defer any decisions. In
addition, other factors militating against prioritizing prevention of
contralateral breast cancer—poor prognosis of the identified cancer,
high risk of other competing causes of morbidity and mortality—
must receive proper consideration.14 For younger women shown to
carry a BRCA1/2 mutation, mastectomies and reconstruction can
then be integrated with systemic therapies, prophylactic salpingooophorectomy, and adjuvant radiation in optimal sequence. The option of delayed contralateral mastectomy for women who may
become aware of their BRCA1/2 status even years after an early age at
initial diagnosis also merits consideration in light of these data, although certainly less important than ovarian cancer prevention, given
the surveillance and hormonal chemoprevention opportunities for
breast cancer. At least equally important for BRCA1/2 carriers more
mature at breast cancer diagnosis, the Graeser et al12 data demonstrate
that their risk of contralateral breast cancer is less compelling. There is
less justification for contralateral prophylactic mastectomy for this
group, and the ordeal of bilateral reconstruction of greater consequence. Future data from the German group may address the lethality
of the second breast cancer and its contribution to mortality in carriers. After all, the goal of the contralateral mastectomy should be the
prevention of mortality from a second tumor, not only the tumor
itself. As the rate of bilateral mastectomy at breast cancer diagnosis in
unselected women is currently rising with remarkable speed in the
United States, the question of what women consider an acceptable risk
© 2009 by American Society of Clinical Oncology
Downloaded from on September 17, 2014. For personal use only. No other uses without permission.
Copyright © 2009 American Society of Clinical Oncology. All rights reserved.
of second breast cancer and how they understand risk, whether or not
they carry a predisposing mutation, must be addressed.15
Oncologists already try to help women to consider a wide range
of probabilities to guide therapeutic decisions: risks and benefits of
various established interventions, risks of treatment-related complications, potential risks and benefits of investigational therapies. The risk
of contralateral breast cancer and its management are already part of
many discussions at initial diagnosis and should be emphasized, but
not overemphasized. As Graeser et al12 have shown, knowledge of
BRCA1/2 mutation status may inform this aspect of the discussion,
providing reassurance to women whose genetic testing is negative and
stratified information to mutation carriers on which to base some
difficult decisions. While the data should further impel us to find
better nonsurgical ways of preventing breast cancer in women at
risk—including breast cancer survivors and women with and without
inherited susceptibilities—for the moment, at least, we can provide
ever more reliable and refined information with which to personalize
our patients’ care.
2. Audeh M, Penson R, Friedlander M, et al: Phase II trial of the oral PARP
inhibitor olaparib (AZD2281) in BRCA-deficient advanced ovarian cancer. J Clin
Oncol 27:274s, 2007 (suppl; abstr 5500)
3. Gronwald J, Byrski T, Huzarski T, et al: Neoadjuvant therapy with cisplatin
in BRCA1-positive breast cancer patients. J Clin Oncol 27:7s, 2009 (suppl; abstr
4. Haffty BG, Harrold E, Khan AJ, et al: Outcome of conservatively managed
early-onset breast cancer by BRCA1/2 status. Lancet 359:1471-1477, 2002
5. Pierce LJ, Levin AM, Rebbeck TR, et al: Ten-year multi-institutional results
of breast-conserving surgery and radiotherapy in BRCA1/2-associated stage I/II
breast cancer. J Clin Oncol 24:2437-2443, 2006
6. Begg CB, Haile RW, Borg A, et al: Variation of breast cancer risk among
BRCA1/2 carriers. JAMA 299:194-201, 2008
7. Metcalfe K, Lynch HT, Ghadirian P, et al: Contralateral breast cancer in
BRCA1 and BRCA2 mutation carriers. J Clin Oncol 22:2328-2335, 2004
8. Schwartz MD, Lerman C, Brogan B, et al: Utilization of BRCA1/BRCA2
mutation testing in newly diagnosed breast cancer patients. Cancer Epidemiol
Biomarkers Prev 14:1003-1007, 2005
9. Tercyak KP, Peshkin BN, Brogan BM, et al: Quality of life after contralateral
prophylactic mastectomy in newly diagnosed high-risk breast cancer patients
who underwent BRCA1/2 gene testing. J Clin Oncol 25:285-291, 2007
10. Palomares MR, Paz B, Weitzel JN: Genetic cancer risk assessment in the
newly diagnosed breast cancer patient is useful and possible in practice. J Clin
Oncol 23:3165-3166, 2005; author reply 3166-3167, 2005
11. Mai PL, Lagos VI, Palomares MR, et al: Contralateral risk-reducing mastectomy in young breast cancer patients with and without genetic cancer risk
assessment. Ann Surg Oncol 15:3415-3421, 2008
12. Graeser M, Engel C, Rhiem K, et al: Contralateral breast cancer risk in
BRCA1 and BRCA2 mutation carriers. J Clin Oncol 27:5887-5892, 2009
13. Gao X, Fisher SG, Emami B: Risk of second primary cancer in the
contralateral breast in women treated for early-stage breast cancer: A populationbased study. Int J Radiat Oncol Biol Phys 56:1038-1045, 2003
14. Recht A: Contralateral prophylactic mastectomy: Caveat emptor. J Clin
Oncol 27:1347-1349, 2009
15. Tuttle TM, Habermann EB, Grund EH, et al: Increasing use of contralateral
prophylactic mastectomy for breast cancer patients: A trend toward more
aggressive surgical treatment. J Clin Oncol 25:5203-5209, 2007
1. Tutt A, Robson M, Garber J, et al: Phase II trial of the oral PARP inhibitor
olaparib in BRCA-deficient advanced breast cancer. J Clin Oncol 27:7s, 2009
(suppl; abstr CRA501)
DOI: 10.1200/JCO.2009.25.1652; published online ahead of print at on October 26, 2009
The author(s) indicated no potential conflicts of interest.
Conception and design: Judy E. Garber
Administrative support: Judy E. Garber
Collection and assembly of data: Judy E. Garber, Mehra Golshan
Data analysis and interpretation: Judy E. Garber, Mehra Golshan
Manuscript writing: Judy E. Garber, Mehra Golshan
Final approval of manuscript: Judy E. Garber, Mehra Golshan
■ ■ ■
Recognition and Treatment of Sleep Disturbances
in Cancer
Sonia Ancoli-Israel, Department of Psychiatry, University of California San Diego, and Rebecca and John Moores University of
California San Diego Cancer Center, San Diego, CA
See accompanying article on page 6033
Fatigue is recognized by oncologists as one of the most frequent
complaints of patients with cancer. More importantly, fatigue is
among the symptoms about which patients express the most concern.
What is less recognized is that there are many components of fatigue,
including physiologic factors (such as pain, anemia or menopause),
psychological factors (such as depression or anxiety), and chronobiologic factors (such as circadian rhythms disorders and sleep).1 In
particular, the relationship between fatigue and sleep is becoming
more clear, with data suggesting that sleep problems are significantly
correlated with increased fatigue.2 Yet, patients with cancer are not
always asked about their sleep nor treated appropriately for their
sleep problems.
Insomnia is defined as difficulty falling asleep, difficulty staying
asleep, and/or nonrestorative sleep, resulting in daytime dysfunction.3
© 2009 by American Society of Clinical Oncology
The most common sleep-related complaints of patients with cancer
are difficulty falling asleep, difficulty staying asleep, and frequent and
prolonged nighttime awakenings.4,5 In other words, patients with
cancer are complaining of insomnia.
The risk factors for insomnia in cancer include the cancer itself
(eg, tumors that increase steroid production, symptoms of tumor
invasion resulting in pain, dyspnea, nausea, pruritus), treatment factors (eg, corticosteroids, hormonal fluctuations), medications (eg,
narcotics, chemotherapy, neuroleptics, sympathomimetics, steroids,
sedative hypnotics), environmental factors (eg, temperature extremes
or too much light or noise in the bedroom), psychosocial disturbances
(eg, depression, anxiety, stress), and comorbid medical disorders (eg,
Downloaded from on September 17, 2014. For personal use only. No other uses without permission.
Copyright © 2009 American Society of Clinical Oncology. All rights reserved.
headaches, other primary sleep disorders).6 In a study of cancer survivors, 52% reported sleeping difficulties, and although two thirds reported their insomnia began before their cancer diagnosis, 58%
reported that having cancer aggravated their sleep problem.7 This
suggests a negative feedback loop where the challenges faced by patients with cancer may contribute to insomnia, which in turn may feed
back to exacerbate medical conditions comorbid with cancer.4 Treatment of the sleep problem at any time point might therefore break
that cycle.
An important aspect of treatment is, of course, identifying the
problem. Sleep needs to be thought of as part of the symptom cluster
often associated with cancer. The concept of symptom clusters is not
new in the field of cancer.8,9 In a study by Liu et al,10 which examined
a symptom cluster of poor sleep, fatigue and depression, results suggested that the more symptoms within that symptom cluster the
patients experienced before the start of chemotherapy, the worse the
symptoms they experienced during chemotherapy. In addition, those
patients with more frequent and more severe symptoms pretreatment
experienced the most severe symptoms during treatment.
However, several studies have shown that many patients with
cancer do not mention their sleep problems, with close to 80% assuming it is caused by the treatment, 60% wrongly assuming that the
symptoms will not last, and almost half believing that their physicians
cannot do anything to help them.11,12 What this means is that clinicians need to include sleep as part of the symptom cluster already
recognized, and to ask all patients about their sleep. Without asking
the question, “How are you sleeping?” this important problem might
never be identified and addressed.
The importance of treatment rises from the knowledge that insomnia results in more severe fatigue, leads to mood disturbances,
contributes to immunosuppression, affects quality of life, and potentially affects the course of the cancer.6,13 The question for every clinician then becomes, “How do I best treat insomnia in my patients
with cancer?”
Insomnia in this patient population may be due to a variety of
causes; therefore, treatment may need to be multimodal and include
both pharmacologic treatment (eg, benzodiazepine receptor agonists
or melatonin receptor agonists) and nonpharmacologic therapies.6,13
The 2005 National Institutes of Health State-of-the-Science Conference statement on insomnia concluded that behavioral therapies
are the most effective treatments for insomnia,3 and there have
now been several studies showing that cognitive behavioral therapy
for insomnia is effective in treating this sleep problem in cancer
survivors.14-17 These studies all confirmed that cognitive behavioral
therapy for insomnia improved sleep efficiency (the percent of time
spent sleeping out of time in bed), increased total sleep time, improved
fatigue and mood (ie, decreased depression and anxiety), and improved quality of life, with therapeutic effects maintained at 3-, 6- and
12-month follow-up.
One of the innovative features of the Berger et al study18 in this
issue of Journal of Clinical Oncology is that intervention was initiated
before the patients with cancer developed sleep disturbances and severe fatigue. Results suggested that although sleep improved at 90 days
postchemotherapy in the group administered behavioral therapy for
insomnia, unlike the studies that initiated treatment postchemotherapy to patients with insomnia, at 1 year there were no longer any
differences between the groups. Whereas Berger et al18 concluded that
clinicians need to identify and intervene with behavioral therapy at the
point that patients with cancer report moderate/severe insomnia, the
other take-home message should be that treatment initiated during
chemotherapy may have short-term benefits, and additional treatment might be needed postchemotherapy. Berger et al18 are correct
that clinicians need to ask their patients about their sleep and initiate
treatment when the problem is identified.
In summary, sleep disorders, particularly insomnia, are common
in patients with cancer. Sleep needs to be assessed carefully in patients
with cancer to improve quality of life and possibly to help improve the
course of the disease. There are a variety of effective pharmacologic
and nonpharmacologic therapies available for the management of
cancer-related insomnia. But for those therapies to work, the clinician
must first identify the problem by communicating with the patient
and then be willing to initiate the appropriate treatment. Only then
will we be able to improve the quality of life for our patients with
cancer during and after their cancer treatment.
Although all authors completed the disclosure declaration, the following
author(s) indicated a financial or other interest that is relevant to the subject
matter under consideration in this article. Certain relationships marked
with a “U” are those for which no compensation was received; those
relationships marked with a “C” were compensated. For a detailed
description of the disclosure categories, or for more information about
ASCO’s conflict of interest policy, please refer to the Author Disclosure
Declaration and the Disclosures of Potential Conflicts of Interest section in
Information for Contributors.
Employment or Leadership Position: None Consultant or Advisory
Role: Sonia Ancoli-Israel, sanofi-aventis (C), Sepracor (C), Ferring
Pharmaceuticals (C), GlaxoSmithKline (C), Orphagen Pharmaceuticals
(C), Pfizer (C), Respironics (C), Schering-Plough (C) Stock Ownership:
None Honoraria: None Research Funding: Sonia Ancoli-Israel,
Sepracor Expert Testimony: None Other Remuneration: None
1. Ancoli-Israel S, Moore P, Jones V: The relationship between fatigue and
sleep in cancer patients: A review. Eur J Cancer Care 10:245-255, 2001
2. Liu L, Ancoli-Israel S: Sleep disturbances in cancer. Psychiatric Annals
38:627-634, 2009
3. National Institutes of Health State of the Science Conference Statement
on Manifestations and Management of Chronic Insomnia in Adults, June 13-15,
2005. Sleep 28:1049-1057, 2005
4. Fiorentino L, Ancoli-Israel S: Insomnia and its treatment in women with
breast cancer. Sleep Med Rev 10:419-429, 2006
5. Engstrom CA, Strohl RA, Rose L, et al: Sleep alterations in cancer patients.
Cancer Nurs 22:143-148, 1999
6. O’Donnell JF: Insomnia in cancer patients. Clin Cornerstone 6:S6-S14,
2004 (suppl 1D)
7. Savard J, Simard S, Blanchet J, et al: Prevalence, clinical characteristics,
and risk factors for insomnia in the context of breast cancer. Sleep 24:583-590,
8. Miaskowski C, Dodd M, Lee K: Symptom clusters: The new frontier in
symptom management research. J Natl Cancer Inst Monogr 17-21, 2004
9. Miller AH, Ancoli-Israel S, Bower JE, et al: Neuroendocrine-immune
mechanisms of behavioral comorbidities in patients with cancer. J Clin Oncol
26:971-982, 2008
10. Liu L, Fiorentino L, Natarajan L, et al: Pretreatment symptom cluster in
breast cancer patients is associated with worse sleep, fatigue and depression
during chemotherapy. Psycho-oncology 18:187-194, 2009
11. Stone P, Richardson A, Ream E, et al: Cancer-related fatigue: Inevitable,
unimportant and untreatable? Results of a multi-centre patient survey—Cancer
Fatigue Forum. Ann Oncol 11:971-975, 2000
12. Curt GA, Breitbart W, Cella D, et al: Impact of cancer-related fatigue on the
lives of patients: New findings from the Fatigue Coalition. Oncologist 5:353-360,
13. Bardwell WA, Profant J, Casden DR, et al: The relative importance of
specific risk factors for insomnia in women treated for early-stage breast cancer.
Psycho-oncology 17:9-18, 2008
© 2009 by American Society of Clinical Oncology
Downloaded from on September 17, 2014. For personal use only. No other uses without permission.
Copyright © 2009 American Society of Clinical Oncology. All rights reserved.
14. Quesnel C, Savard J, Simard S, et al: Efficacy of cognitive-behavioral therapy
for insomnia in women treated for nonmetastatic breast cancer. J Consult Clin
Psychol 71:189-200, 2003
15. Savard J, Simard S, Ivers H, et al: Randomized study on the efficacy of
cognitive-behavioral therapy for insomnia secondary to breast cancer, part I:
Sleep and psychological effects. J Clin Oncol 23:6083-6096, 2005
16. Espie CA, Fleming L, Cassidy J, et al: Randomized controlled clinical effectiveness trial of cognitive behavior therapy compared with treatment as usual for
persistent insomnia in patients with cancer. J Clin Oncol 26:4651-4658, 2008
17. Fiorentino L, McQuaid JR, Liu L, et al: Cognitive behavioral therapy for
insomnia in breast cancer survivors: A randomized controlled crossover study.
Sleep 31:A295, 2008
18. Berger AM, Kuhn BR, Farr LA, et al: One-year outcomes of a behavioral
therapy intervention trial on sleep quality and cancer-related fatigue. J Clin Oncol
27:6033-6041, 2009
DOI: 10.1200/JCO.2009.24.5993; published online ahead of print at on November 2, 2009
■ ■ ■
The Forest and the Trees: Pathways and Proteins As
Colorectal Cancer Biomarkers
Monica M. Bertagnolli, Brigham and Women’s Hospital, Dana Farber Cancer Institute, Boston, MA
See accompanying articles on pages 5924 and 5931
In a 1990 review, Fearon and Vogelstein1 presented a model for
the genetic basis of colorectal neoplasia, stating that colorectal cancer
(CRC) development requires the accumulation of mutations in multiple genes that regulate cell growth and differentiation. They proposed that “identification of the genetic alterations present in tumors
may provide a molecular tool for improved estimation of prognosis in
patients with CRC . . . multiple pathways exist in which new chemotherapeutic agents might achieve a therapeutic advantage.”1(p764) The
molecular characteristics described in the 1990 Fearon and Vogelstein
review included mutational activation of the oncogenes c-myc and
KRAS and tumor suppressor loss by mutation of TP53 or allelic loss
at chromosome 18q. These events occur at a relatively high frequency in CRC; yet, almost two decades later, we still have much to
learn concerning the prognostic or predictive value of these four
markers, and that of the many other tumor-associated characteristics
subsequently identified.
This issue of Journal of Clinical Oncology includes two articles
concerning K-Ras,2,3 a protein whose inactivation in CRC was first
observed in 1987 but has only recently been identified as a significant
clinical biomarker.4,5 K-Ras activation occurs downstream of epidermal growth factor receptor (EGFR), and studies of CRCs from
patients treated with the anti-EGFR antibodies panitumumab or
cetuximab showed that mutational activation of KRAS predicts lack
of treatment response. These studies involved both retrospective
tissue collections from non–randomly assigned patients and correlative studies from prospectively randomized clinical trials of antiEGFR therapy. The results were striking, showing that responses to
anti-EGFR– containing regimens were equal to controls for patients
with K-Ras mutant tumors. Differences in progression-free survival
for antibody-treated patients whose tumors were with or without
KRAS mutations were on the order of 2 to 5 months, in favor of the
wild-type cases (reviewed in Walther et al6).
Laurent-Puig et al2 retrospectively studied 173 advanced CRC
cases collected from six hospitals, of which all but one received a
cetuximab-containing regimen as second-line or greater therapy.
They examined additional members of the EGFR signaling pathway,
predicting that KRAS wild-type tumors would fail to respond to cetuximab if signaling was driven by other mechanisms of constitutive
pathway activation. Consistent with known regulatory mechanisms of
© 2009 by American Society of Clinical Oncology
EGFR signaling, they found that EGFR amplification predicted improved cetuximab response. In addition, activation of pathway members K-Ras or BRAF, or loss of the phosptastase and tensin homolog
tumor suppressor, correlated with lack of clinical response. If these
results are confirmed in additional studies, then as many as 70% of
patients with metastatic CRC may reasonably be excluded from
EGFR-directed therapies. In addition, analysis of other pathway members, such as PI3K (PIK3CA), may further improve the ability to
predict anti-EGFR response. It is anticipated that these results will also
hold for use of anti-EGFR agents in the adjuvant setting. Collectively,
the clinical correlation of tumor EGFR pathway activation status and
targeted agent response represents a major advance, sparing the majority of patients with advanced CRC therapies that are both costly
and ineffective.
It is still not clear whether constitutive activation of EGFR pathway is
in itself a negative prognostic factor for CRC. One crude way of assessing
this is to examine the prevalence of these signaling changes across the
different clinical stages of CRC. Microsatellite instability (MSI), the best
understood colon cancer molecular prognostic factor, is present in
roughly 25% to 30% of stage II, 15% to 20% of stage III, and less than 10%
of stage IV disease, consistent with its characterization in many clinical
biomarker analyses as a predictor of less aggressive behavior. This same
approach suggests that the presence of a KRAS mutation is probably not
prognostic, as the prevalence of K-Ras activation is approximately 35% to
55% across all cancer stages, with the higher value achieved by testing for
multiple uncommon KRAS mutations. The existing prognostic data concerning K-Ras involve small studies indicating that K-Ras-mutant tumors
carry a worse prognosis, and a few larger studies reporting no association
with outcome (reviewed in ref 6). A second report in this issue, from
Richman et al, provides data using prospectively collected tissues from the
Medical Research Council Fluorouracil, Oxaliplatin and Irinotecan: Use
and Sequencing (FOCUS) trial, a large study of advanced CRC patients
that was conducted from 2000 to 2003. Patients included in this biomarkeranalysiswererandomlyassignedtoreceiveeitherfirst-linefluorouracil
(FU), followed by either FU/irinotecan or FU/oxaliplatin on progression,
or FU/irinotecan or FU/oxaliplatin as first-line therapy, with no protocolspecified second-line treatment. This group tested tumors from 711 patients for mutations in KRAS and BRAF. They found that the presence of
these mutations predicted poor overall survival, but no difference in
Downloaded from on September 17, 2014. For personal use only. No other uses without permission.
Copyright © 2009 American Society of Clinical Oncology. All rights reserved.
disease-free survival. Unfortunately, despite the high quality of this study,
anti-EGFR therapy was available for CRC clinical trials in Europe
during the enrollment period of the MRC FOCUS trial, raising the
possibility that second line treatment could have biased this result.
A search of PubMed for “colorectal cancer biomarker” yields over
11,500 entries as of August, 2009. The majority of these publications
report positive results, yet despite so many possibilities, KRAS is the first
biomarker to achieve significant clinical utility. The reasons for this are
clear. First and foremost, CRC is a highly heterogeneous disease. At the
time of the Fearon and Vogelstein review,1 few anticipated the degree of
molecular heterogeneity that powerful new tools to interrogate the CRC
genome would uncover. In 2006 and 2007, sequence data from 20,857
transcripts representing 18,191 distinct genes were examined for 11 CRCs
to identify genes that were mutated in a tumor but not in normal tissue
from the same patient.7,8 An additional 96 CRCs were then used to determine the prevalence of changes found in the initial tumor set. This work
revealed an astounding degree of heterogeneity. Each CRC contained an
average of 15 mutations likely to contribute to tumor behavior, but very
few of these defects were common among the different tumors.8 In fact,
most of the cancer-associated genes were mutated in less than 5% of the
cancers studied. Above and beyond this genomic heterogeneity are layers
of complexity introduced by variations in other determinants of tumor
biology, such as gene and protein expression, immune response, and
epithelial-stromal interactions.
Unfortunately, current clinical trials resources do not allow easy
correlation of uncommon tumor molecular characteristics with treatmentresponseoroutcome.Themajorityofbiomarkerstudiesreportdata
from relatively small, retrospective tissue collections, and it is likely that
this literature is limited by a publication bias toward positive studies.
Biomarkers identified in this manner will fail to proceed to validation and
clinical use unless the marker is present at a relatively high prevalence and
indicates a significant change in clinical behavior. Although higher quality
data are emerging from cancer clinical trials, the issue of sample size
remains a major hurdle for even the largest adjuvant treatment study.
Additional noise is introduced into the analysis when biomarkers are
studied in patient cohorts that have not been randomly assigned to treatment; in some cases, random assignment based on the biomarker itself is
required. In the optimal study design, the biomarker is used to assign
treatment. This design is currently used in E5202, a study of adjuvant
therapy for patients with stage II colon cancer where chemotherapy is
administered only to those with non-MSI tumors that also demonstrate 18qLOH.
The field of biomarker development is substantially aided by development of targeted therapeutics. The EGFR/K-Ras story is an excellent
example of how understanding the signaling pathways involved in clinical
response to a targeted agent leads to successful identification of patients
most likely to benefit from treatment. As demonstrated by the two reports
in this issue, once a pathway member is implicated, then the remainder
can be tested, and even events of low overall frequency can be understood
in clinical context. Clinical validation of a biomarker also guides new drug
discovery. In this case, using methods such as synthetic screens for lethality, researchers are searching for agents that will overcome constitutive
activation of K-Ras. These data will provide further insight into the biology of EGFR, and likely lead to new effective biomarkers and treatments
for the K-Ras-mutant tumor subset.
One of the principles set forth in the 1990 Fearon and Vogelstein
review1 was that it takes multiple molecular changes to produce CRC.
It is likely, therefore, that multiple markers will be required to adequately distinguish clinical behavior. For CRC, some testable patterns
have already emerged. Examples of this are the subdivision of disease
according to tumor-specific mechanisms of genomic instability, including MSI, CpG island methylator phenotype, and chromosomal
instability. In addition, pattern analyses from large-scale genotyping
and gene expression studies have found tumor-specific disruption of
pathways that govern cell adhesion, inflammatory response, the cytoskeleton, and the extracellular matrix. The relevance of these exciting new insights cannot be understood without access to human
clinical resources, because animal models and tumor xenografts are
inadequate systems for understanding the relationship between
these complex multifactor disease characteristics and clinical behavior. The most promising way to move the biomarker field forward
is to continue to develop our understanding of the interrelated signaling networks that govern tumor biology, using well-annotated tissue
collections obtained in the setting of cancer clinical trials.
A 2007 summary of progress by Wood et al8 concluded that “the
compendium of genetic changes in individual tumors provides new opportunities for individualized diagnosis and treatment of cancer. Taking
advantage of these opportunities is the major challenge for the future.”(p1108) Understanding the molecular basis of CRC has turned out to
be a lot more complicated and difficult than anticipated, and so has the
search for clinically useful biomarkers. The lesson from both past failures
and current successes is that these challenges can be met only through
dedicated collaboration between basic and clinical cancer research.
Although all authors completed the disclosure declaration, the following
author(s) indicated a financial or other interest that is relevant to the subject
matter under consideration in this article. Certain relationships marked with a
“U” are those for which no compensation was received; those relationships
marked with a “C” were compensated. For a detailed description of the
disclosure categories, or for more information about ASCO’s conflict of interest
policy, please refer to the Author Disclosure Declaration and the Disclosures of
Potential Conflicts of Interest section in Information for Contributors.
Employment or Leadership Position: None Consultant or Advisory
Role: None Stock Ownership: None Honoraria: Monica M. Bertagnolli,
Pfizer Research Funding: None Expert Testimony: None Other
Remuneration: None
1. Fearon ER, Vogelstein B: A genetic model for colorectal tumorigenesis. Cell
61:759-767, 1990
2. Laurent-Puig G, Cayre A, Manceau G, et al: Analysis of PTEN, BRAF, and
EGFR status in determining benefit from cetuximab therapy in wild-type KRAS
metastatic colorectal cancer. J Clin Oncol 27:5924-5930, 2009
3. Richman SD, Seymour MT, Chambers P, et al: KRAS and BRAF mutations
in advanced colorectal cancer are associated with poor prognosis but do not
preclude benefit from oxaliplatin or irinotecan: Results from the MRC FOCUS
trial. J Clin Oncol 27:5931-5937, 2009
4. Bos JL, Fearon ER, Hamilton SR, et al: Prevalence of ras gene mutations in
human colorectal cancers. Nature 327:293-297, 1987
5. Forrester K, Almoguera C, Han K, et al: Detection of high incidence of K-ras
oncogenes during human colon tumorigenesis. Nature 327:298-303, 1987
6. Walther A, Johnston E, Swanton C, et al: Genetic prognostic and predictive
markers in colorectal cancer. Nat Rev Cancer 9:489-499, 2009
7. Sjoblom T, Jones S, Wood LD, et al: The consensus coding sequences of
human breast and colorectal cancers. Science 314:268-274, 2006
8. Wood LD, Parsons DW, Jones S, et al: The genomic landscapes of human
breast and colorectal cancers. Science 318:1108-1113, 2007
DOI: 10.1200/JCO.2009.24.8013; published online ahead of print at on November 2, 2009
■ ■ ■
© 2009 by American Society of Clinical Oncology
Downloaded from on September 17, 2014. For personal use only. No other uses without permission.
Copyright © 2009 American Society of Clinical Oncology. All rights reserved.
Journal Corrections
The June 20, 2008, article by Albers et al, entitled, “Randomized Phase III Trial Comparing Retroperitoneal Lymph
Node Dissection With One Course of Bleomycin and Etoposide
Plus Cisplatin Chemotherapy in the Adjuvant Treatment of
Clinical Stage I Nonseminomatous Testicular Germ Cell Tumors:
AUO Trial AH 01/94 by the German Testicular Cancer Study
Group” (J Clin Oncol 26:2966-2972, 2008), contained an error.
In the Patients and Methods section, under Treatment and
Follow-Up, Arm A, the dose for bleomycin was given as 30,000
U, whereas it should have been 30,000 IU (30 mg), as follows:
“One cycle of BEP chemotherapy was administered using
following dosages: cisplatin 20 mg/m2, days 1 to 5, 60-minute
infusion; etoposide 100 mg/m2, days 1 to 5, 60-minute infusion;
bleomycin 30,000 IU (30 mg), days 1, 8, and 15, bolus infusion.”
The online version has been corrected in departure from
the print. Journal of Clinical Oncology apologizes to the authors
and readers for the mistake.
DOI: 10.1200/JCO.2010.28.7417
■ ■ ■
The December 10, 2009, editorial by DeAngelis, entitled,
“Anaplastic Gliomas” (J Clin Oncol 27:5861-5862, 2009), contained an error.
The title should have been, “Anaplastic Glioma: How to
Prognosticate Outcome and Choose a Treatment Strategy.”
The online version has been corrected in departure from
the print. Journal of Clinical Oncology apologizes to the author
and readers for the mistake.
DOI: 10.1200/JCO.2010.28.7425
© 2010 by American Society of Clinical Oncology