close

Вход

Забыли?

вход по аккаунту

?

559

код для вставкиСкачать
1197
CORRESPONDENCE
The Concept of Tenju-gann, or “Natural-End
Cancer”
he concept of Tenju-gann proposed by Kitagawa et al.1 and discussed in the accompanying editorial by Kennedy2 deserves
broader application than these authors suggest.
Although cancer is increasingly an illness of the elderly, there
remains a significant burden of disease among the young and middleaged, particularly in terms of years of potential life lost. There is also
evidence that some solid tumors occurring in those younger than 40
years behave more aggressively and result in poorer outcomes than
the same disease diagnosed at a more advanced age.3,4 Being cheated
of health and forced to pay the cancer “tax” before enjoying the fruits
of a long life is a devastating situation that leads patients, families,
and physicians to pursue every reasonable attempt at therapy. The
vast majority of advanced solid cancers become refractory to therapy
over time or never demonstrate initial sensitivity to systemic treatment. In many cases, symptomatic and supportive care become the
best and most appropriate therapeutic options. In these situations,
the concept of Tenju-gann or “natural-end cancer,” as presented by
Kitagawa et al.3 could be applied to patients of all ages, including
children, in an effort to help them and their families cope with the
shattering consequences of terminal disease. The increasing focus on
quality of life issues in clinical trial design as well as the development
of hospice organizations are practical attempts at understanding and
relieving the suffering of advanced malignancy. The concept of Tenjugann may further aid in our care of individuals with terminal disease,
regardless of age. It could help avoid the application of medically
futile therapeutic procedures that interfere with the ability of our
patients to experience the “peaceful death with minimal suffering”
that is central to the diagnosis of a Tenju-gann cancer. This does not
imply therapeutic nihilism, but rather the admission that, for most
patients diagnosed with advanced cancer, succumbing to their disease is inevitable and leads to a “natural-end” despite aggressive
therapeutic efforts. Dying of advanced cancer at age 40 years is no less
“natural” than dying of the same illness at age 90 years.
Oncologists can play a critical role in helping patients of any age
accept their disease and experience a peaceful death. Perhaps there is
a derivative of Tenju that could encompass those dying after having
enjoyed shorter-than-expected lives. We should not let reverse ageism
deprive our younger patients of the potential comfort that is the
ultimate result of the application of the philosophy of Tenju-gann.
T
REFERENCES
1.
2.
3.
4.
© 1999 American Cancer Society
Kitagawa T, Hara M, Sano T, Sugimura T. The concept of Tenju-gann, or
“natural-end cancer.” Cancer 1998;83:1061– 65.
Kennedy BJ. Cancer death and older age. Cancer 1998;83:1066 – 8.
Nixon AJ, Neuberg D, Hayes DF, Gelman R, Connolly JL, Schnitt S, et al.
Relationship of patient age to pathologic features of the tumor and prognosis
for patients with stage I or II breast cancer. J Clin Oncol 1994;12:888 –94.
Winchester DP, Osteen RT, Menck HR. The National Cancer Data Base report
1198
CANCER March 1, 1999 / Volume 85 / Number 5
on breast carcinoma characteristics and outcome in relation
to age. Cancer 1996;78:1838 – 43.
3.
Kennedy BJ. Communicating with patients about advanced
cancer [letter to the editor]. JAMA 1998;280:1403– 4.
Daniel Rayson, M.D.
Department of Oncology
University of Western Ontario
London, Ontario, Canada
B. J. Kennedy, M.D.
Regents’ Professor Emeritus of Medicine
Masonic Professor Emeritus of Oncology
University of Minnesota
Minneapolis, Minnesota
Author Reply
Author Reply
C
W
ancer at any age is an unwanted burden. Quality of
life is an essential part of cancer management,
regardless of age. The article by Kitigawa et al. on
Tenju-gann emphasizes a peaceful death from cancer
for older persons.
Dr. Rayson emphasizes that young persons with
uncured cancer are deprived of the pleasures of a long
life. Clinical cancer research has concentrated on patients younger than 70 years. For some, this research
has paid off. Acute leukemia in children appears cured
in up to 80% of cases, advanced testicular carcinoma
is curable in over 90% of males (mainly in those ages
17–36 years), and cervical carcinoma in young women
is curable—the result of early detection efforts.
The biologic behavior of cancer differs with age.
Dr. Rayson correctly points out that young women
with breast carcinoma have more aggressive disease
than older women. However, in Hodgkin disease,
stage for stage, older persons have worse outcomes
than younger ones.1
With the worldwide increase in longevity (Japan
has had the greatest), the problem of cancer increases;
over 50% of cancers occur in persons older than 67
years. There has been a paucity of education and
research regarding this older patient group. The recognition of this deficiency has stimulated efforts to
encourage programs in geriatric oncology.2 Emphasis
on the quality of life of an older patient and appropriate selection of therapies can be regarded as an extension of what has been emphasized in caring for cancer
patients of young age.
Dr. Rayson refers to patients with “terminal disease.” The word terminal with respect to cancer
should be deleted from the oncologist’s vocabulary.
Reference to the end-of-life phase of cancer is preferable.3
REFERENCES
1.
2.
Kennedy BJ, Fremgren AM, Menck HR. Aging and Hodgkin’s
disease. Ann Oncol 1998;9(Suppl 3):75.
Kennedy BJ. Aging and cancer: geriatric oncology. Cancer
1997;80:1270 –2.
e thank Dr. Rayson for expressing interest in the
concept of Tenju-gann (pronounced “ten-JEWgan”) and welcome the opportunity to respond to his
comments.
Dr. Rayson suggests that the concept of Tenjugann deserves broader application to diseases among
the young and the middle-aged, because the vast majority of advanced solid cancers in these age groups
are refractory to therapy. It is naturally very conceivable that the Tenju-gann mind-set might be of great
help in reconciling patients and family members to
terminal disease, regardless of age, and assist in treatment decisions to allow a “peaceful death with minimal suffering.” We should make every effort to minimize a cancer patient’s physical, psychologic, and
social burdens. The more widespread development of
caring and affordable hospice organizations is urgently needed, and emphasis on palliative care is essential. In these matters, we completely agree with Dr.
Rayson. However, whereas the mind-set of Tenjugann applies to all age groups, the concept as defined
can be applied only to those in old age. We would like
to explain our thoughts further to clarify a different
nuance of the issue. This will be done first from a
Japanese perspective and then from a wider perspective.
The Japanese Perspective
Tenju in Japanese means “healthy, long life,” which is,
in traditional Japanese thought, given by heaven to
every person at birth. With this concept as the background, we can celebrate when a person has experienced Tenju without encountering or having to deal
with serious diseases or accidents. Within the boundaries of this term, a line should be drawn between
mere fatalism and the Tenju concept.
Most importantly, care should be taken that the
Tenju-gann concept not be exploited by therapeutic
nihilists. In Japan, in recent years, a book entitled
Patients, Don’t Fight Against Your Cancer!, written
by a radiologist, became particularly popular. The
author correctly pointed out the problematic atti-
Correspondence
tude of some surgeons and medical oncologists who
perform aggressive treatments in the simplistic subjective belief that those treatments are beneficial,
without either objective proof or concern about the
patient’s quality of life. However, the strong and to
some extent inaccurate statements expressed in the
book, combined with repeated indiscriminate coverage in the press, has caused a worrisome nihilism
among people in their attitudes toward therapy. A
tendency now exists among some cancer patients to
avoid not only appropriate medical treatments, but
also even health checks for early detection of cancer. The backlash clearly has had negative consequences for many cancer patients and for Japanese
society in general.
The Broader International Perspective
Many people feel that cancers, including those occurring in patients of young age, are inevitable.
However, there have been many situations in which
various types of cancer became preventable and
stopped occurring. Others are becoming preventable and declining in their occurrence. Examples of
both are occupational cancers, gastric carcinomas,
hepatocellular carcinomas after hepatitis B or C virus infection, and squamous cell carcinomas in
smokers. Thus, we should be generally optimistic
about preventing, or at least delaying, the onset of
cancer. Possible cancer development in people with
hereditary predisposition could also be eventually
avoided or lessened by sophisticated genetic counseling or improved treatment in earlier phases of the
disease.
Moreover, advanced techniques for diagnosing
cancers at an early stage and improved accessibility to
those techniques should significantly improve the rate
of permanent cure. Subjects treated successfully for
primary cancers could develop second and third primary cancers, but recurrence could be well managed.
It is no longer rare to see complete cure in cases of
multiple cancers.
According to Buddhist teaching, heaven gave a
life span of about 120 years to humans. L. A. Loeb
calculated that prostate carcinoma would develop at
the late age of 130 years if the mutation rate could
be diminished to half the current rate. Of course,
this provides a reason for general optimism, but it
needs to be supported by efficacy and progress in
cancer prevention as well as improved early diagnosis and treatment. Within the broadest definition of
“treatment,” we include mental and psychologic
1199
treatment in a manner that allows us to reach the
goal of Tenju-gann.
Tomoyuki Kitagawa, M.D.
The Cancer Institute of the Japanese Foundation for
Cancer Research
Tokyo, Japan
Takashi Sugimura, M.D.
National Cancer Center
Tokyo, Japan
Primary Staging and Follow-Up of
High Risk Melanoma Patients with
Whole-Body 18F-Fluorodeoxyglucose
Positron Emission Tomography
Results of a Prospective Study of 100
Patients
e read with interest the article by Rinne et al.,1
who prospectively evaluated 100 consecutive
high risk melanoma patients with whole-body
positron emission tomography (PET) and conventional diagnostics. In 52 cases (Group A), PET was
performed for primary staging. In 48 cases (Group B),
there was clinical suspicion of progression, and therefore PET was performed for restaging during the follow-up. The authors concluded that PET “is a highly
sensitive and specific technique for melanoma staging, with exception of the brain.” This study is particularly important because it is the first large prospective study to apply PET in the primary staging of
melanoma patients. The authors define “high risk
melanoma” as a tumor thickness of at least 1.5 mm
and Clark Level IV. However, the mean Breslow thickness of the primary lesions is not mentioned, nor is
the range. What is “high risk”? This discussion dates
from the 1970s, when the term “high risk” was introduced.2 An elective lymph node dissection, performed
on melanoma patients with Breslow thickness 1.5– 4.0
mm, has shown to confer little, if any, survival advantage. The risk of metastases varies strongly according
to thickness, and we think the term “high risk” is
relative and needs reevaluation based on current
knowledge, particularly regarding lymphatic dissemination.
We agree with the authors that the earliest possible detection of metastases at a potentially curable
stage is a major objective. PET appears to be very
W
1200
CANCER March, 1 1999 / Volume 85 / Number 5
useful in Group B. However, in Group A, PET detected
lymph node metastases in only 8% of patients,
whereas no systemic metastases were found. With this
percentage taken into account, we wonder if PET is
sensitive enough for accurate regional staging. Biopsy
of the first tumor-draining, sentinel lymph node (SN)
is the most recent developed method for the detection
of early micrometastases in the regional lymph
nodes.3 SN biopsy, which was not mentioned in the
article by Rinne et al., has the potential to become the
procedure of choice in the regional staging of melanoma patients. With the SN biopsy, we found lymph
node metastases in 28% and 67% of Stage I melanoma
patients with Breslow thickness 1.51– 4.0 and .4.0
mm, respectively,4 and this finding was confirmed by
many others. It is not imaginary that a significant
number of small metastases were missed by PET. Unfortunately, the sizes of the nine lymph node metastases detected by PET were not reported in the article.
Borgstein et al. studied the size and distribution of
tumor metastases within 50 tumor positive SN.5 They
discovered that with an SN biopsy, micrometastases
as small as 0.0005 mm2 could be detected. From subsequent lymph node dissections it became clear that
in 75% of cases the SN was the only tumor positive
lymph node, with a median metastasis area of 0.15
mm2. It is unclear, but very unlikely, that micrometastases smaller than 1 mm2 can be detected by 18Ffluorodeoxyglucose PET. Practical support for this is
found in a recent study by Macfarlane et al.6
In summary, we think that SN biopsy is to be
preferable for initial regional staging of Stage I melanoma. The SN biopsy has a lower detection limit (theoretically, one cell) and no risk of false-positivity, and
it is obtainable in more hospitals than PET. In our
opinion, the diagnostic accuracy of both SN biopsy
and PET imaging needs to be prospectively studied in
a large group of (unselected) patients with a higher
prevalence of lymphatic metastases at presentation
(Breslow thickness .3.0 or 4.0 mm). In a broader
sense, the still-unknown prognostic relevance of detection of micrometastatic lymph node involvement
could affect the place of either technique in the clinical management algorithm.
REFERENCES
1.
2.
3.
Rinne D, Baum RP, Hör G, Kaufmann R. Primary staging and
follow-up of high risk melanoma patients with whole-body
18
F-fluorodeoxyglucose positron emission tomography.
Cancer 1998;82:1664 –71.
Balch CM, Murad TM, Soong SJ, Ingalls AL, Richards PC,
Maddox WA. Tumor thickness as a guide for surgical management of clinical stage I melanoma patients. Cancer 1979;
43:883– 8.
Joseph E, Messina J, Glass FL, Cruse CW, Rapaport DP,
4.
5.
6.
Berman C, et al. Radioguided surgery for the ultrastaging of
the patient with melanoma. Cancer J Sci Am 1997;3:341–5.
Pijpers R, Borgstein PJ, Meijer S, Hoekstra OS, van Hattum LH,
Teule GJJ. Sentinel node biopsy in melanoma patients: dynamic lymphoscintigraphy followed by intraoperative gamma
probe and vital dye guidance. World J Surg 1997;21:788 –93.
Borgstein P, Pijpers R, van Diest P, Meijer S. Staging early
lymphatic metastases in melanoma: the case for super-selective lymph node dissection [poster presentation]. SSO Cancer
Symposium, San Diego, California, March 26–29, 1998.
Macfarlane DJ, Sondak V, Johnson T, Wahl RL. Prospective
evaluation of 2-[18F]-2-deoxy-D-glucose positron emission
tomography in staging of regional lymph nodes in patients
with cutaneous malignant melanoma. J Clin Oncol 1998;16:
1770 – 6.
G. Sophie Mijnhout, M.D.
Rik Pijpers, M.D.
Otto S. Hoekstra, M.D., PhD.
G. Jaap J. Teule, M.D., PhD.
Department of Nuclear Medicine
Academisch Ziekenhuis van de Vrije Universiteit
Amsterdam, The Netherlands
Paul J. Borgstein, M.D.
Sybren Meijer, M.D., PhD.
Department of Surgical Oncology
Academisch Ziekenhuis van de Vrije Universiteit
Amsterdam, The Netherlands
Author Reply
W
e are pleased to respond to the comments of Mijnhout et al. with respect to our article1 on primary
staging and follow-up of high risk melanoma patients
with 18F-fluorodeoxyglucose positron emission tomography (PET). According to these authors, sentinel lymph
node biopsy should be preferred for the initial regional
staging of Stage I melanoma because it has a lower
detection limit and no risk of false positivity.
In our study, melanoma patients with a tumor
thickness .1.5 mm (range, 1.5–10.0 mm; mean, 2.75
mm) were investigated because it is commonly accepted that thicker melanomas have a higher risk of
metastasizing than thinner ones (and therefore staging procedures are usually more extensive for lesions
.1.5 mm). We are aware that there is not sharp cutoff
point in risk groups, but rather a continuous increase
of potential tumor spread with advanced thickness. In
melanoma .4 mm, initial systemic dissemination is
more likely. In the context of our study, however, the
discussion of risk groups was less relevant.
Primary staging of patients with thicker melanomas should exclude distant metastases, which are
usually a contraindication for aggressive surgical exci-
Correspondence
sion.2 Here our study showed a superiority of PET over
conventional techniques, which was confirmed by
many others.2,3 Wagner et al.3 reported PET scans of
12 patients undergoing dissections of 14 lymph node
basins. PET scans predicted presence and absence of
metastases correctly in all lymph nodes. Similar results were reported by others.2– 4 Due to nonmorphologic imaging, PET cannot measure lymph node size.
Lesions smaller than slice thickness (3.4 mm) can be
missed, and it is most unlikely that initial tumor nests
in sentinel lymph nodes would be detectable with
current PET technology. It is remarkable that in our
study all nine lymph node metastases were detected
by PET and missed by conventional imaging. Certainly, accuracy of sentinel lymph node biopsy is currently best for early detection of regional lymph node
involvement, but information regarding distant
spread cannot be obtained. Mijnhout et al. stated that
elective lymph node dissection performed on patients
with melanomas of Breslow thickness 1.5– 4.0 mm has
shown little, if any, survival advantage. It still has to be
determined whether an early lymph node dissection
following positive sentinel lymph node biopsy confers
any survival benefit on the patient. PET combines a
high accuracy in lymph node staging with an additional whole-body examination. The aim of our study
was exclusively to investigate the value of PET compared with conventional diagnostics in primary staging and follow-up, but we share the opinion that comparison of PET and sentinel lymph node biopsy needs
further assessment.
REFERENCES
1.
2.
3.
4.
Rinne D, Baum RP, Hör G, Kaufmann R. Primary staging and
follow-up of high risk melanoma patients with whole-body
18
F-fluorodeoxyglucose positron emission tomography.
Cancer 1998;82:1664 –71.
Holder W, White RL, Zuger JH, Easton EJ, Greene FL. Effectiveness of positron emission tomography for the detection
of melanoma metastases. Ann Surg 1998;227:764 –71.
Wagner JD, Schauwecker D, Hutchins G, Coleman JJ. Initial
assessment of positron emission tomography for detection
of nonpalpable regional lymphatic metastases in melanoma. J Surg Oncol 1997;64:181–9.
Macfarlane DJ, Sondak V, Johnson T, Wahl RL. Prospective
evaluation of 2- [18 F]-2-Deoxy-D-Glucose positron emission tomography in staging of regional lymph nodes in
patients with cutaneous malignant melanoma. J Clin Oncol
1998;16:1770 – 6.
Prediction of Axillary Lymph Node
Involvement of Women with
Invasive Breast Carcinoma
A Multivariate Analysis
hile reading the article by Olivotto et al.,1 I tried
to interpret their Table 5 according to its caption,
its footnote, its contents, and the referring text. I soon
began to wonder whether the table consisted of descriptive statistics rather than predictive probabilities.
If my surmise is correct, then even the 77% accuracy of
the authors’ 3-variable model would not justify, to me,
using Table 5 to estimate risk, let alone to define risk
groups.
As a monitory example, can one be absolutely
certain on viewing Table 5 (“observed probability” 5
0.00 for 2 patients) that a woman with a 0.5-cm tumor,
lymphvascular invasion, and no palpable disease will
not have axillary lymph node involvement?
In either case, if a table is all a clinician needs
(sans parameters and equations), the ordinary Bayesian2,3 multivariate analysis approach can be applied to
the plentiful data of Olivotto et al. (their Table 1,
.4500 patients), with the results shown in my table
(Table 1).
The format of my Table 1 is the same as that of
Table 5 by Olivotto et al. for ease of comparison. One
sees that the Bayesian probability is a nonmonotonic
function of tumor size and that it has a maximum for
tumor sizes between 3.1–5.0 cm.
The Bayesian probability corresponding to the example given earlier is 36%.
W
REFERENCES
1.
2.
3.
Daniela Rinne, M.D.
Roland Kaufmann, M.D.
Department of Dermatology
University of Frankfurt Medical Center
Frankfurt, Germany
1201
Olivotto IA, Jackson JSH, Mates D, Andersen S, Davidson W,
Bryce CJ, et al. Prediction of axillary lymph node involvement of women with invasive breast carcinoma. Cancer
1998;83:948 –55.
Bayes T. An essay towards solving a problem in the doctrine
of chances. The Philosophical Transactions 1763 [posthumous];53:370 – 418. Reproduced in Biometrika 1958;45:293–
315.
Box GEP, Tiao GC. Bayesian inference in statistical analysis.
New York: John Wiley & Sons, Inc., 1992.
Donald I. Promish, M.S.
Burlington, Vermont
1202
CANCER March, 1 1999 / Volume 85 / Number 5
TABLE 1
Probability of Axillary Lymph Node Involvement of Women with Invasive Breast Carcinoma as a Function of Tumor Size, Palpability, and
Lymphvascular Invasion
Size of tumor (cm)
Palpability
LVI absent
Nonpalpable
Palpable primary
Axillary lymph
nodes, palpable
LVI present
Nonpalpable
Palpable primary
Axillary lymph
nodes, palpable
0–0.5
0.6–1.0
1.1–1.5
1.6–2.0
2.1–2.5
2.6–3.0
3.1–5.0
5.1–10.0
0.11
0.21
0.05
0.10
0.07
0.14
0.09
0.18
0.16
0.30
0.27
0.45
0.29
0.47
0.15
0.28
0.55
0.33
0.43
0.50
0.66
0.79
0.80
0.64
0.36
0.56
0.19
0.34
0.26
0.44
0.31
0.50
0.48
0.67
0.63
0.79
0.65
0.81
0.45
0.64
0.85
0.70
0.78
0.82
0.90
0.95
0.95
0.89
LVI: lymphvascular invasion.
Author Reply
U
nfortunately, there are significant flaws in Mr.
Promish’s interpretation of our results and his
Bayesian estimation procedure. First, the data contained in Table 5 of our article1 are observed (descriptive) not estimated (predicted) probabilities of lymph
node metastases. To use Mr. Promish’s monitory example, Table 5 shows that neither of the 2 women in
our sample with # 0.5-cm, nonpalpable tumors with
lymphvascular invasion (LVI) had lymph node metastases at presentation. However, we should point out
that whether the data contained in Table 5 are observed or estimated probabilities, the practicing clinician should never presume to be “absolutely certain”
that a particular patient has any specific probability of
lymph node metastases based on the results of published retrospective analyses. This is especially true for
women with a relatively unusual or rare pattern of
prognostic characteristics such as # 0.5 cm, nonpalpable tumors with LVI. This simply is because one can
never be absolutely certain that the prognostic characteristics or features of a particular patient are exactly
equivalent to the characteristics or features of the
sample on which the analyses were conducted. Furthermore, the practicing clinician also should keep in
mind that any statistic has an associated confidence
interval within which one reasonably can expect the
corresponding population parameter to vary. The
confidence interval on a proportion involving two patients is very wide.
Mr. Promish also implies that his Bayesian estimation procedure may give a better representation of
the risk of lymph node metastases than our logistic
regression model. When evaluating the performance
of any estimation procedure, the practicing clinician
always should consider carefully the agreement between the actual, observed results and the results estimated or predicted by the estimation procedure.
When disagreement is large, it is wise to question
the validity of the procedure used to generate the
estimates. One should be concerned about the validity of the Bayesian procedure used by Mr. Promish
simply because it generated estimates that differ
markedly from the actual probabilities reported in
our Table 5. In contrast, our three variable logistic
regression model-generated estimates that differ very
little from the actual probabilities given in Table 5.
Furthermore, unlike the Bayesian estimates, the
monotone relation between tumor size and probability of lymph node metastases that has been observed
by other authors2–5 clearly is evident in the estimates generated by our logistic regression procedure.
We recently cross-validated our 3-variable model on a
second, independent sample of 3108 women and
found nearly identical probabilities of lymph node
metastases for the risk groups shown in Table 5 in our
article.1
REFERENCES
1.
2.
3.
Olivotto IA, Jackson JSH, Mates D, Andersen S, Davidson W,
Bryce CJ, et al. Prediction of axillary lymph node involvement of women with invasive breast carcinoma. Cancer
1998;83:948 –55.
Barth A, Craig PH, Silverstein MJ. Predictors of axillary
lymph node metastases in patients with T1 breast carcinoma. Cancer 1997;79:1918 –22.
De Laurenetis M, Gallo C, De Placido S, Perrone F, Pettianato G, Petrella G, et al. A predictive index of axillary nodal
involvement in operable breast cancer. Br J Cancer 1996;73:
1241–7.
Correspondence
4.
5.
Choong PL, deSilva CJS, Dawkins HJS, Sterret GF, Robbins P,
Harvey JM, et al. Predicting axillary lymph node metastases
in breast carcinoma patients. Breast Cancer Res Treat 1996;
97:135– 49.
Carter CL, Allen C, Henson D. Relation of tumor size, lymph
node status and survival in 24,740 breast cancer cases. Cancer 1989;63:181–7.
Jeremy S. H. Jackson, Ph.D.
Ivo A. Olivotto, M.D.
Donna Mates, M.A.
British Columbia Cancer Agency
Vancouver and Fraser Valley Clinics
Vancouver and Surrey, British Columbia
Canada
cycles of chemotherapy. At 32 weeks, a cesarean section was performed and immediately was followed by
a radical abdominal hysterectomy with pelvic and
paraaortic lymphadenectomy. Pediatric follow-up of
the newborn showed no sign of any metabolic or
hematologic abnormality. The auditory brain stem
evoked potential test was normal and at last follow-up
the child was of normal weight, size, and cognitive and
psychomotor development.
REFERENCES
1.
2.
Neoadjuvant Chemotherapy in the
Treatment of Locally Advanced
Cervical Carcinoma in Pregnancy
3.
n a recent article Tewari et al. reported two cases of
neoadjuvant chemotherapy in the treatment of locally advanced cervical carcinoma in pregnancy.1 In
the abstract conclusion, they stated that to their
knowledge their report constituted the first description of the use of chemotherapy in such patients
(MEDLINE, 1966 –1997). Although it was a useful review regarding chemotherapy during pregnancy, we
disagree with their conclusion. The earliest report is
that of Jacobs et al.2 in 1981 for an oat cell carcinoma
of the cervix. Therapy was comprised of a single injection (50 mg/kg) of cisplatin at 10 week’s amenorrhea, followed by a radical hysterectomy 2 weeks later.
Therefore this report is of little assistance with respect
to tumor response and the effect of chemotherapy on
the fetus. However, in 1996 we published a very similar report of one case with cisplatin-based neoadjuvant chemotherapy in a woman who refused to interrupt her pregnancy.3 Treatment was comprised of
single agent therapy with cisplatin for 3 cycles of 75
mg per m2 body area. The treatments were administered at 22, 25, and 28 completed weeks of gestation.
There was a complete clinical response after the three
Tewari K, Cappuccini F, Gambino A, Kohler MF, Pecorelli S,
DiSaia PJ. Neoadjuvant chemotherapy in the treatment of
locally advanced cervical carcinoma in pregnancy: a report
of two cases and review of issues specific to the management of cervical carcinoma in pregnancy including planned
delay of therapy. Cancer 1998;82:1529 –34.
Jacobs AJ, Marchevsky A, Gordon RE, Deppe G, Cohen CJ.
Oat cell carcinoma of the uterine cervix in a pregnant
woman treated with cis-diamminedichloroplatinium. Gynecol Oncol 1980;9:405–10.
Giacalone PL, Laffargue F, Benos P, Rousseau O, Hedon B.
Cisplatinium neoadjuvant chemotherapy in a pregnant
woman with invasive carcinoma of the uterine cervix. Br J
Obstet Gynaecol 1996;103:932– 4.
Pierre-Ludovic Giacalone, M.D.
François Laffargue, M.D.
Department of Obstetrics and Gynecology
Hôpital Arnaud de Villeneuve
Montpellier, France
A Report of Two Cases and Review of
Issues Specific to the Management of
Cervical Carcinoma in Pregnancy
Including Planned Delay of Therapy
I
1203
Author Reply
T
he comments by Drs. Giacalone and Laffargue are
appreciated. However, two points need to be addressed. Drs. Giacolone and Laffargue state that in our
abstract conclusion we claim that our report constitutes the first description of the use of neoadjuvant
chemotherapy for cervical carcinoma in pregnancy,
but if the abstract conclusion is read carefully, we
stated that to our knowledge, these cases constituted
the first description of the use of neoadjuvant chemotherapy for invasive squamous cell carcinoma of the
cervix in pregnancy.1 We are fully aware of the report
by Jacobs et al. in 1980, but these investigators were
describing the management of a neuroendocrine or
oat cell cervical carcinoma, not a squamous cell cancer.2 (In point of fact, we recently published an account of neuroendocrine carcinoma in pregnancy3
and have included the reference of Jacobs et al.)
Second, our article was prepared during the first
half of 1997 and submitted for consideration for pub-
1204
CANCER March, 1 1999 / Volume 85 / Number 5
lication in Cancer in June 1997. The article written by
Giacalone et al. appeared in September 19964 but was
not referenced through MEDLINE at the time we were
preparing our article. We occasionally have noted a
delay in the entry of case reports and research letters
to the MEDLINE database. As an example, one of our
articles was published in the February 1998 issue of
The Lancet,5 but did not appear in the MEDLINE database until October 1998.
We acknolwedge the significance of the report
by Giacalone et al. The second patient in our article
was treated in 1995 and therefore it is reasonable to
assume that both groups (ours and Giacalone et al.)
treated similar patients with neoadjuvant chemotherapy during the same period. It is our opinion
that both reports are noteworthy contributions and
attest to the importance of this treatment option for
women with locally advanced disease who refuse to
interrupt their pregnancies. Because we have an
interest in the management of cancer in pregnancy,
we look forward to further communications on this
subject.
REFERENCES
1.
2.
3.
4.
5.
Tewari K, Cappuccini F, Gambino A, Kohler MF, Pecorelli
S, DiSaia PJ. Neoadjuvant chemotherapy in the treatment
of locally advanced cervical carcinoma in pregnancy: a
report of two cases and review of issues specific to
the management of cervical carcinoma in pregnancy including planned delay of therapy. Cancer 1998;82:1529 –
34.
Jacobs AJ, Marchevsky A, Gordon RE, Deppe G, Cohen CJ.
Oat cell carcinoma of the uterine cervix in a pregnant
woman treated with cis-diamminedichloroplatinum. Gynecol Oncol 1980;9:405–10.
Balderston KD, Tewari K, Gregory WT, Berman ML, Kucera
PR. Neuroendocrine small cell uterine cervix cancer in pregnancy. Gynecol Oncol 1998;71:128 –32.
Giacalone P-L, Laffargue F, Benos P, Rousseau O, Hedon B.
Cis-platinum neoadjuvant chemotherapy in a pregnant
woman with invasive carcinoma of the uterine cervix. Br J
Obstet Gynaecol 1996;103:932– 4.
Tewari K, Rose GS, Balderston KD, Porto M, Kohler MF.
Fertility drugs and malignant germ cell tumor of the ovary in
pregnancy. Lancet 1998;351:957– 8.
Krishnansu Tewari, M.D.
Philip J. DiSaia, M.D.
Division of Gynecologic Oncology
University of California, Irvine
Medical Center
Orange, California
Ten-Year Disease Free Survival
after Transperineal SonographyGuided Iodine-125 Brachytherapy
with or without 45-Gray External
Beam Irradiation in the Treatment
of Patients with Clinically
Localized, Low to High Gleason
Grade Prostate Carcinoma
I
n their article, Ragde et al. employ classic techniques of misinformation in their attempt to show
that implants have better results than external beam
irradiation. Their group of patients, who have favorable prognostic features, are compared with my series of patients of all stages and grades. The reference they selected to quote is an older one; it
demonstrated that conformal treatment produced
higher 5-year bNED rates than conventional treatment. I will not comment on the inappropriate use
of Stamey’s work as representative of radiation therapy or defend other institutions whose work has
been similarly distorted.
Table 1 compares the data of Ragde et al. with
various endpoints for the 1995 reference quoted and a
more appropriate 1997 publication of our work. Our
1997 data show that 61% of all patients are bNED at 5
years compared with 74% and 71% for Ragde et al.
This is remarkably good, considering that twice as
many (57% vs. 29%) of our patients had prostate specific antigen (PSA) levels .10 ng/mL and one-fourth
(9% vs. 38%) as many had favorable ,4 ng/mL levels.
PSA is the strongest independent predictor of failure.
Our series had nearly twice as many T1 patients (30%
vs. 17%) but 6 times as many T3 patients. When external beam irradiated patients with favorable tumor
characteristics are compared with the favorable or
total implant group, 5-year results with the external
beam irradiation are actually numerically higher, as
shown in the table.
The statements by Ragde et al. and subsequent
newspaper quotations about external beam irradiation are not accurate and have caused many patients
needless worry about the treatment they have received or are considering. An improved or equally
effective new treatment can stand on its own merits
and does not need to be justified by a distorted interpretation of the findings of other studies.
Correspondence
1205
TABLE 1
Comparison of Endpoints for Three Studies of External Beam Irradiation in the Treatment of Patients with Prostate Carcinoma
Patient characteristics
Hanks et al. 19952
Hanks et al. 19973
Ragde et al. 19981
Radiotherapy
No. of patients
T1
T2
T3
Gleason score
2–6
7–10
PSA
,15
.15
0–10
10–20
201
5-yr bNED
All
Conformal and conventional
502
14%
66%
20%
Conformal and conventional
456
30%
59%
12%
152
17%
81%
2%
2–7
8–10
,4 ng/mL
4–,10 ng/mL
.10 ng/mL
96%
4%
9%
34%
57%
,4 ng/mL
4–10 ng/mL
.10 ng/mL
91%
9%
38%
33%
29%
All
61%
All
Favorable
74%
71%
77%
23%
60%
40%
44% (50% conformal)
(39% conventional)
T1–2, all grades
PSA
,4
4–10
10–20
201
89%
83%
66%
28%
PSA: prostate specific antigen. bNED: no biochemical or clinical evidence of disease.
REFERENCES
1.
2.
3.
Ragde H, Elgamal AA, Snow PB, Brandt J, Bartolucci AA,
Nadir BS, Korb LJ. Ten-year disease free survival after transperineal sonography-guided iodine-125 brachytherapy with
or without 45-Gray external beam irradiation in the treatment of patients with clinically localized, low to high Gleason grade prostate carcinoma. Cancer 1998;83:989 –1001.
Hanks GE, Lee WR, Schultheiss TE. Clinical and biochemical
evidence of control of prostate cancer at 5 years after external beam radiation. J Urol 1995;154:456 –9.
Hanks GE, Hanlon AL, Schultheiss TE, Freedman GM, Hunt
M, Pinover WH, Mousas B. Conformal external beam treatment of prostate cancer. Urology 1997;50:87–92.
Gerald E. Hanks, M.D.
Department of Radiation Oncology
Fox Chase Cancer Center
Philadelphia, Pennsylvania
Author Reply
W
e thank Dr. Hanks for his comments and the
opportunity to amplify further the differences
between our study and the most recent study cited by
him. It is our belief that we do not need to use “techniques of misinformation” to show that results are
better with implants than with external beam irradiation; these findings will stand up to any scrutiny. However, our choice of studies for comparison was based
primarily on an attempt to provide similar periods of
follow-up. The Fox Chase Cancer Center (FCCC) 1995
study was chosen because it had the longest follow-up
(median, 50 months) from a regional center of excellence that we felt should be included. The FCCC study
of 1997 (median, 40 months) did not have a comparable follow-up period.
First, let us state that the patient populations
studied by the two groups are indeed different, as
noted by Dr. Hanks. However, we believe the major
difference is not as easily explained by merely noting
percentages of groups with certain prostate specific
antigen (PSA) levels. Our study population had their
treatments completed by June 1988, a time when PSA
levels and their implications were not well understood. Many patients who were placed into our “low
risk” group by virtue of their Gleason score and/or
clinical stage actually had PSA levels far greater than
10 ng/mL. Many patients treated by Dr. Ragde were
1206
CANCER March, 1 1999 / Volume 85 / Number 5
initially diagnosed with one or more prostate nodules
and not by the more recent innovation of PSA screening, ergo the relatively low number of T1 patients in
our series (17%) compared with that of Dr. Hanks
(30%), and hence it follows that our study had the
higher number of palpable lesions (83% vs. 70%). To
address the issue of numbers of T3 patients, we will
say only that the notorious variability in clinical staging makes comparison between percentages of T2 and
T3 groups pointless.
Secondly, there is the issue of Gleason scores.
There has been a documented phenomenon of
Gleason score progression over the last several
years. Previously reported in the early years of prostate brachytherapy, Gleason scores of 2 and 3 are
now extinct and scores of 4 and 5 are on the endangered species list. This trend, coupled with the possibility of institutional variability and potential
overscoring of pathology specimens, may invite
skepticism when Dr. Hank’s study of patients
treated before 1993 is compared with our current
study of patients treated 5 years earlier. However,
this too may be a moot point, because our study also
has more poorly differentiated tumors (9%) when
compared with the 1997 article cited (4%).
Finally, we have a serious concern regarding the
wisdom of comparing patients with different biologic
and statistical endpoints. Dr. Hanks does indeed
present admirable results for patients treated with
external beam irradiation. However, it is difficult for
us to accept that his study, using endpoints (2 consecutive PSA rises equal to 1.5 ng/mL) at 7 years evaluated
with a Kaplan–Meier actuarial analysis, is able to
stand side by side with the current study, which used
10-year observed disease free survival with an endpoint of PSA less than or equal to 0.5.
Perhaps a better comparison can be made between our 5-year overall bNED reported earlier,1
which used a level of 1.0 ng/mL for the definition of
PSA failure. However, even given this more stringent
PSA cutoff than the 1.5 ng/mL used in the FCCC
1997 study, we reported a 87% bNED at 5 years,
compared with 61% reported by FCCC. However, we
will admit that external beam irradiation alone may
be a good treatment choice for some patients; indeed, external beam irradiation is already proving
itself to be a valuable adjunct to seeding in high risk
patients.
We do thank Dr. Hanks for his very thought-provoking comments. It is the continued open discussion
of issues like these that will lead to a greater understanding of this disease by all.
REFERENCES
1.
Ragde H. Interstitial iodine-125 radiation without adjuvant
therapy in the treatment of clinically localized prostate carcinoma. Cancer 1997;80:442–53.
Haakon Ragde, M.D.
Leroy Korb, M.D.
Abdel-Aziz Elgamal, M.D.
Northwest Prostate Institute
Seattle, Washington
I
n 1982, as a young and impressionable physicianin-training looking for possible employment, I visited Dr. John Blasko in Seattle. He happened to have a
patient scheduled for a prostate implant and was kind
enough to take me into the operating room with him.
His template-based transperineal technique was radically different from the free hand suprapubic approach in vogue at the University of California and
elsewhere at the time, and I was duly impressed. Ever
since, I have followed Dr. Blasko’s work with a keen
interest. It was therefore a great delight to read another excellent article on this subject from the Seattle
area.1
With a much larger number of patients treated
using the technique reported in that article, many of
the conclusions reached by Ragde et al. may well
prove to be true in the future. However, in the meantime, it should be pointed out that even though it is
certainly impressive to have actual (as opposed to
actuarial) 10-year data on this many patients, one still
has to keep in mind that the conclusions are based on
only 152 nonrandomly treated patients. One would
also prefer to know more details regarding the treatment-related morbidity before accepting some of the
conclusions on combined external beam irradiation
and brachytherapy used routinely to treat selected
patients. Finally, one could quibble with the authors’
use of the terms “less invasive” and “less intrusive” (in
the third paragraph on page 990 and the first paragraph on page 998, respectively) in reference to prostate implants. It is likely that when the authors used
those terms, they were comparing brachytherapy with
prostatectomy. However, one hastens to point out that
before the verdict is in, external beam irradiation
alone is still the least invasive modality.
Finally, lest the young and impressionable physicians-in-training become too easily enamored with
the artificial neural network methodology, one has to
be mindful, at all times, that the quality of the answer
provided by any analysis depends mainly on the quality of the data input. In this particular instance, the
study was not set up to compare brachytherapy alone
with brachytherapy plus periprostatic external irradi-
Correspondence
ation. And above all, this now not-so-young and notso-impressionable practitioner likes to put things in
perspective for physicians-in-training regarding a disease that is managed with a whole range of alternatives (from no initial treatment to various combinations of surgical, radiotherapeutic, and hormonal
options) by paraphrasing a rather well-known British
politician: “Never in the history of cancer treatment
have there been so much data accumulated with so
few practical conclusions reached as in the management of patients with prostate cancer.” This, by the
way, is precisely why the Internet is inundated with
advice offered by various pundits on the subject.
REFERENCE
1.
Ragde H, Elgamal A, Snow P, Brandt J, Bartolucci A, Nadir
BS, et al. Ten-year disease free survival after transperineal
sonography-guided iodine-125 brachytherapy with or without 45-Gray external beam irradiation in the treatment of
patients with clinically localized, low to high Gleason grade
prostate carcinoma. Cancer 1998;83:989 –1001.
K. C. Lee, M.D.
Department of Radiation Medicine
Georgetown University Medical Center
Washington, DC
Author Reply
W
e thank Dr. Lee for taking the time to comment
on our article. The letter, however, contains several misconceptions. Contrary to what he believes he
saw in Seattle in 1982, in actuality the first-ever ultrasound-directed transperineal prostate implant in the
U.S. was performed by the senior author of our article,
Dr. Haakon Ragde, on November 20, 1985, at Seattle’s
Northwest Hospital. Perhaps Dr. Lee was watching Dr.
Blasko perform the now-outdated open abdominal
implantation of the prostate, which is no longer widely
used because of its poor tumor control and unwanted
side effects.
With regard to the much-improved transperineal
technique, Dr. Ragde has performed all the implants,
on patients now numbering close to 3000, at Northwest Hospital, with the assistance of several capable
radiation oncologists. Moreover, for Dr. Lee to refer to
the implant procedure as “his [Dr. Blasko’s] templatebased technique” is yet another incorrect statement.
Dr. Blasko would be the first to tell him that the
technique used at Northwest Hospital was originally
developed by Dr. Hans Holm, Professor of Urology at
the University of Copenhagen, Denmark,1 and
brought to Seattle by Dr. Ragde in late 1984.
1207
We disagree with Dr. Lee’s assertion that 152 patients may be too small a number on which to base
definitive conclusions. Much of current medical and
surgical oncology practice is based on the results of
studies that reported on the results of markedly fewer
patients. However, what matters here is that the study
has been done, showing for the first time that the long
term efficacy of prostate brachytherapy is no longer a
matter of speculation, but a matter of record. Furthermore, evaluation of patients having undergone prostate brachytherapy at Northwest Hospital is an ongoing research program, and we are currently preparing
for publication another article that contains a significantly larger number of patients.
We are fully aware of the shortcomings of any
retrospective study and have attempted to address
many of these in our discussion. However, we doubt
that the obvious solution—a prospective, randomized
study dealing with this particular disease modality—
will be done anytime soon in the U.S.
Dr. Lee does correctly point out that treatmentrelated morbidity was notably absent in the article.
Most of the study population had already been evaluated for complications and previously reported. Similar brachytherapy morbidity results are now being
reported by a number of other institutions.2– 4 Repetition of these data was not felt to add substantially to
the article’s merit.
A pronouncement of Dr. Lee’s that was apparently
intended to demonstrate the weak points of computer
networking, i.e., that the quality of input determines
the quality of output (as preferable as that expression
may be to alternate euphemisms), was also misapplied. Certain confusions about the ends and the
means of neural networking may be present. A neural
network evaluation, with all its cutting-edge newness,
is the ideal tool for analyzing retrospectively the variables available in our study. The XAIM team who
performed the analysis has many years of experience
in both medical (including prostate carcinoma) and
military applications and is considered an international leader in this type of technology application.
We do again thank Dr. Lee for the opportunity to
clarify some of the misperceptions that not only he, but
also others, may have had about the history of prostate
brachytherapy and the current work being done by Dr.
Ragde and his team at Northwest Hospital.
REFERENCES
1.
Holm HH, Juul N, Pedersen JF, Hansen H, Stroyer I. Transperineal 125-iodine seed implantation in prostatic cancer guided by
transrectal ultrasonography. J Urol 1983;130:283–6.
1208
2.
3.
4.
CANCER March, 1 1999 / Volume 85 / Number 5
Joly F, Brune D, Couette JE, Lesaunier F, Heron JF, Peny J,
Henry-Amar M. Health-related quality of life and sequelae in
patients treated with brachytherapy and external beam irradiation for localized prostate cancer. Ann Oncol 1998 Jul;9:751–7.
Terk MD, Stock RG, Stone NN. Identification of patients at
increased risk for prolonged urinary retention following radioactive seed implantation of the prostate. J Urol 1988;160:
1379 – 82.
Desai J, Stock RG, Stone NN, Iannuzzi C, DeWyngaert JK.
Acute urinary morbidity following I-125 interstitial implantation of the prostate gland. Radiat Oncol Investig 1988;6:
135– 41.
Haakon Ragde, M.D.
Leroy Korb, M.D.
Northwest Prostate Institute
Seattle, Washington
The Greatest Dimension of Prostate
Carcinoma Is a Simple, Inexpensive
Predictor of Prostate Specific
Antigen Failure in Radical
Prostatectomy Specimens
I
n a recent issue of Cancer, Renshaw et al. reported
on the greatest dimension of prostate carcinoma in
prostatectomy specimens as a potential predictor of
prostate specific antigen (PSA) failure.1 Significance of
the research question that the authors addressed is
indisputable. Unfortunately, some of the main results
were presented in a manner that was not quite
straightforward.
The authors used univariate and multivariate
analyses to evaluate the ability of each of the measured factors (independent variables) to predict time
to postoperative PSA failure (dependent variable). Because the time to PSA failure (continuous variable)
rather than whether each patient experienced PSA
failure in a fixed time period (binary variable) was
used, I assume proportional hazards models were
used. Some of the independent variables were included in statistical models as continuous variables
(e.g., greatest tumor dimension in cm) whereas others
were included as categorical variables (e.g., Gleason
score . 7 vs. Gleason score # 7). Table 2 of the article
by Renshaw et al. shows the “risk ratio” estimate and
associated P value for each of the seven independent
variables based on univariate analyses. What is confusing are the reported “risk ratio” estimates for continuous variables. For example, the risk ratio estimate
for “(tumor) area” was 1.007 (P value 5 0.0011). The
reader who is less literate in biostatistics may wonder
why the risk ratio so close to unity could give such a
highly significant P value, especially given a small
sample size of the study. The reader with more background in biostatistics would guess that the authors
must have calculated the increase in risk of PSA failure
for a unit increase in the independent variable being
analyzed. For a continuous independent variable, the
risk ratio (strictly speaking, hazard ratio if a proportional hazards model was used, but this does not affect
my argument here) between two individuals who differ by the quantity D on the i-th independent variable
(xi) and are the same for all other independent variables is represented by the quantity exp (biD), in which
bi is a regression coefficient for the i-th independent
variable from the proportional hazards model.2 Alternatively, the hazard ratio also can be interpreted as
the instantaneous relative risk of an event per unit
time for an individual with risk factor level xi 1 D
compared with an individual with risk factor level xi,
given that both individuals have survived to time t.
Therefore, the interpretation of the above cited result
will be that the risk of PSA failure increased by 0.7%
with an increase in “(tumor) area” by 1 cm2. Accordingly, the risk ratio of the patient with the largest
tumor area (3.7 cm2) compared with the patient with
the smallest tumor area (0.06 cm2) within the study
population is estimated as exp[ln(1.007) 3 (3.7– 0.06)]
TABLE 1
Reconstruction of the Data Regarding Continuous Variables from Table 2 of Renshaw et al.1
Variable (unit)
P value
Reported risk ratio
bia
Min
Max
Risk ratio for Max vs. Minb
Greatest dimension (cm)
Area (cm2)
Sum of areas (cm2)
Serum PSA (ng/mL)
0.0001
0.0011
0.0002
0.0001
1.206
1.007
1.004
1.086
0.187
0.00698
0.00399
0.0825
0.3
0.06
0.1
2.8
2.5
3.7
7.6
39.1
1.51
1.03
1.03
19.98
Min: minimum value observed; Max: maximum value observed; PSA: prostate specific antigen.
a
Natural logarithm of the reported risk ratio.
b
Exp[bi 3 (Max 2 Min)].
Correspondence
5 1.03, in which bi is calculated as the natural logarithm of the reported “risk ratio.” This means that
the risk of PSA failure was 3% higher for the patient with the largest tumor area when compared
with the patient with the smallest tumor area. I performed similar calculations for the other continuous
variables (Table 1). I used maximum and minimum
values for each variable for the sake of making my
point, but alternative reference values could be
chosen. I believe this format of data presentation,
with clear specification of comparisons being made,
would facilitate the reader’s understanding and critical reading.
P values are useful in sifting data for associations that are more likely to be real or worth pursuing. However, they do not communicate the magnitude of the association. Presenting the results of
statistical analyses in a reader-friendly manner is
essential to putting them in a context to which the
reader can relate. Data from clinical investigation,
such as that by Renshaw et al., would be more
appreciated if the reader could interpret them without having to convert the results into a different
format to evaluate their implications. Clearly, my
critique does not concern the article by Renshaw et
al. alone. I happened to read the article because of
my interest in its subject matter, and wondered
whether busy urologists would have time to do the
calculation shown in Table 1 here, even if they knew
how. Given the emphasis of Cancer on reporting
clinical and epidemiologic data, for which proportional hazards models and logistic regression models frequently are used, I hope the authors and
editors try to leave as little of the burden of making
sense of data as possible with the reader.
REFERENCES
1.
2.
Renshaw AA, Richie JP, Loughlin KR, Jiroutek M, Chung A,
D’Amico AV. The greatest dimension of prostate carcinoma
is a simple, inexpensive predictor of prostate specific antigen failure in radical prostatectomy specimens. Cancer
1998;83:748 –52.
Rosner B. Fundamentals of biostatistics. 3rd edition. Boston:
PWS-Kent Publishing Company, 1989:438 – 40.
Atsuko Shibata, M.D., Ph.D.
Department of Health
Research and Policy
Stanford University
School of Medicine
Stanford, Californiajla
1209
Author Reply
W
e thank Dr. Shibata for taking an interest in our
recent article.1 Our goal was to promote the reporting of estimates of tumor size in radical prostatectomy specimens, and we are heartened to read that
the “significance of the research question . . . is indisputable.”
We also agree with Dr. Shibata that risk ratios for
continuous variables can be difficult to interpret because they represent the risk associated with very
small incremental changes that may not reflect the
way that continuous variables are used in practice. As
a result it is not uncommon, as is the case in our
report, for the risk ratios associated with continuous
variables to appear smaller than for those associated
with categoric variables, even though the continuous
variables may be associated with the same or more
overall risk in a particular data set.
Nevertheless, we do not agree with Dr. Shibata’s
solution to this problem. Although reporting the risk
ratio for maximum versus minimum values in a data
set may appear to reflect more accurately the true risk
present within a particular data set, the results of this
analysis as set forth in Dr. Shibata’s Table 1 are not
any more accurate than those we have presented.
Indeed, for greatest dimension, area, and sum of areas, the reported risk ratios simply are not very different from those that we have reported. More important, the enormous risk ratio of 19.98 reported for
serum prostate specific antigen (PSA) is extremely
misleading and likely reflects the effect of a single
outlier. Serum PSA was not associated with a 20-fold
increase in risk of PSA failure over and above that
associated with tumor size. Finally, Dr. Shibata’s
method does not make sense for categoric data, and if
one were to use both categoric and continuous variables, one would be forced to use two different methods of reporting risk ratios, which may add to the
readers’ confusion.
Thus, we agree that risk ratios associated with
continuous variables can be “confusing,” especially
for “the reader who is less literate in biostatistics,” and
we certainly wish to strive “to leave as little of the
burden of making sense of the data as possible with
the reader.” However, data such as ours are presented
routinely in clinical journals such as Cancer, most
commonly in the format that we have used. If a more
“reader-friendly” and accurate method were to be described and universally accepted we would be among
the first to adopt it; we do not have any particular
attachment to the current format. Unfortunately, Dr.
Shibata’s method is not that method; the results are
1210
CANCER March 1, 1999 / Volume 85 / Number 5
susceptible to bias by outlying data points, and when
not biased, simply are not that different from the data
we already have presented. Until such a method becomes available, and frankly, even when it does become available, a minimum level of statistical sophistication will be required of readers who are interested
in these studies.
REFERENCE
1.
Renshaw AA, Richie JP, Loughlin KR, Jiroutek M, Chung A,
D’Amico AV. The greatest dimension of prostate carcinoma
is a simple, inexpensive predictor of prostate specific antigen failure in radical prostatectomy specimens. Cancer
1998;83:748 –52.
Andrew Renshaw, M.D.
Department of Pathology
Baptist Hospital of Miami
Miami, Florida
Michael Jiroutek, M.S.
University of North Carolina
School of Public Health
Department of Biostatistics
Chapel Hill, North Carolina
Performance Status and
Comorbidity in Elderly Cancer
Patients Compared with Young
Patients with Neoplasia and Elderly
Patients without Neoplastic
Conditions
A
s geriatricians we are clinically and scientifically
involved in the topic of comorbidity. Geriatrics
may even be defined as the medical speciality of comorbidity and functional performance in frail elderly
patients. Therefore, we were very much interested in
the study of Repetto et al., in which the authors promised in title and abstract that they had compared performance and comorbidity in elderly cancer patients
with young patients with neoplasia and elderly patients without neoplastic conditions.1 However, the
authors completely fail to fulfill their promise.
We have two major objections to the methods that
were applied. First, the control group of 278 patients
with a mean age of 75 years in fact is not a nonneoplastic group admitted because of “intercurrent acute
illnesses or reactivation of chronic comorbid conditions” as stated by the authors. In Table 3, in which the
distribution of subjects by cancer stage is presented,
the subjects in this group also turn out to be cancer
patients, with 147 patients with breast carcinoma, 89
patients with colon carcinoma and 79 with prostate
carcinoma. This suggests that there even was a considerable number of subjects in this control group
with two or more neoplasms. How should we explain
this? Is it a major mistake in the data presented in this
table, or is the group in fact a neoplastic patient group,
which was not considered eligible to be referred to
special care centers for cancer treatment. If this last
hypothesis is true, the title is misleading. Moreover, it
is obvious that one will find more comorbidity in such
a group because there had to be at least one more
comorbid condition for the patient to be admitted to
the hospital and be included in the study.
Second, the study lacks a definition of comorbidity, which turns out to be a crucial omission. Instead,
the authors have selected a number of conditions,
without explaining how they came to this selection.
This results in a list of chronic conditions in which
acute (e.g., fractures) and chronic diseases (e.g., deafness), psychiatric and physical diseases, and chronic
diseases with significant impact (e.g., diabetes mellitus) and minor impact on functional performance
(e.g., kyphosis) are compared. By simply adding the
number of conditions found in each patient, the authors conclude that elderly cancer patients have a
higher prevalence of comorbidity compared with
younger patients, but a lower prevalence compared
with older noncancer patients.
There already are a number of properly validated
indices of comorbidity available that do correct for the
severity of the comorbid conditions.2,3 It would have
been much better if one of these had been applied in
this study.
The only thing the study of Repetto et al. clearly
shows is that scientific research into comorbidity in
geriatric patients is difficult to perform in a methodologically correct way. Correct subject selection and
definition of comorbidity are crucial in the design of
future studies in this field. For the moment, it appears
more promising to focus on the results of comorbidity
by measuring functional limitations with well defined
scales of activities of daily living and instrumental
activities of daily living.4,5 Most likely, these measures
will be more helpful in guiding clinical decision making in geriatric patients with malignancies.
REFERENCES
1.
Repetto L, Venturino A, Vercelli M, Gianni W, Biancardi V,
Casella C, et al. Performance status and comorbidity in
elderly cancer patients compared with young patients with
neoplasia and elderly patients without neoplastic conditions. Cancer 1998;82:760 –5.
Correspondence
2.
3.
4.
5.
Incalzi RA, Capparella O, Gemma A, Landi F, Bruno E, Di
Meo F, et al. The interaction between age and comorbidity
contributes to predicting the mortality of geriatric patients
in the acute-care hospital. J Intern Med 1997;242:291– 8.
Charlson M, Szatrowski TP, Peterson J, Gold J. Validation of
a combined comorbidity index. J Clin Epidemiol 1994;47:
1245–51.
Stafford RS, Cyr P. The impact of cancer on the physical
function of the elderly and their utilization of health care.
Cancer 1997;80:1973– 80.
Fratino L, Serraino D, Zagonel V. The impact of cancer on
the physical function of the elderly and their utilization of
health care [letter]. Cancer 1998;83:589 –91.
Marcel G. M. Olde Rikkert, M.D., Ph.D.
Annemie M. W. J. Diepstraten, M.D.
Department of Geriatric Medicine
University Hospital Nijmegen
Nijmegen, The Netherlands
Author Reply
W
ith regard to the first major objection of Drs.
Rikkert and Diepstraten to the method applied in
our article, there is a mistake in the data presented in
Table 3. In our study there were only 2 groups of
neoplastic patients: those age . and those age , 70
years (Group A and Group B, respectively). In the third
column of Table 3, erroneously reported as Group C
rather than the total, we reported the distribution by
stage of all neoplastic patients. When performing comorbidity analyses, conditions prompting the initial
hospital admission were excluded (i.e., only comorbid
disease in addition to the condition prompting the
initial hospital admission were considered).
With regard to the second objection to our study
(lack of definition of comorbidities), we investigated
the presence or absence of comorbidity based on a list
of 25 different conditions chosen for their high prevalence. Our results regarding the type and the rank of
conditions are similar to those reported by other authors as stated in the “Discussion” section. The presence of a comorbidity was based on established definitions, reviewed medical records, and information
collected by standardized interviews for confirmation.
When measuring comorbidity we face a number
of different conditions, with different degrees of severity, impact on function, and combinations. Data re-
1211
garding the number of conditions are too crude and
have little clinical implication. In our study we demonstrated that older cancer patients referred to a cancer center for specialized care present with lower comorbidity compared with patients of similar age
admitted to departments of medicine for intercurrent
illnesses. Several instruments for measuring comorbidities have been proposed, but to our knowledge
none has been validated in elderly cancer patients.
With regard to the proposal to focus on the measurement of functional limitations to better analyze
comorbidity, there are two studies1,2 that have investigated comorbidities and functioning measured by
Comprehensive Geriatric Assessment in elderly cancer
patients that conclude that these variables largely are
independent. In our study comorbidity was measured
by the number of conditions.1 In the study by Exterman et al.2 comorbidity was measured by the Charlson comorbidity index and Cumulative Illness Rating
Scale for Geriatrics.
Evaluating older patients is a difficult and timeconsuming process. Future research should focus on
identifying the most effective and largely applicable instruments to measure both comorbidity and function.
REFERENCES
1.
2.
Repetto L, Venturino A, Comandini D, Fratino L, Gianni W,
Audisio RA, Zagonel V. The role of comprehensive geriatric
assessment in the elderly cancer patients. Ann Oncol 1998;
9(3):52.
Extermann M, Overcash J, Lyman GH, Parr J, Balducci L.
Comorbidity and functional status are independent in older
cancer patients. J Clin Oncol 1998;16:1582–7.
Lazzaro Repetto, M.D., Ph.D.
Divisione di Oncologia Medica I
Istituto Nazionale per la Ricerca
sul Cancro
Genova, Italy
Walter Gianni, M.D.
Isituto Geriatria
Universita La Sapienza
Rome, Italy
Danila Comandini, M.D.
Divisione di Oncologia Medica I
Istituto Nazionale per la Ricerca
sul Cancro
Genova, Italy
Документ
Категория
Без категории
Просмотров
2
Размер файла
125 Кб
Теги
559
1/--страниц
Пожаловаться на содержимое документа