Gunnar Steineck's Group

Denna sida på svenska

The Division of Clinical Cancer Epidemiology is a continuation of the research group Professor Gunnar Steineck set up after his dissertation in 1990. In recent years many of our activities have been focussed on long-term follow-up of cancer survivors. The Division has developed new clinical epidemiological methods for documentation of symptoms, today the health care system can aim higher than before with an ambition to cure without causing harm. Another high-priority project is long-term follow-up of the relatives of patients who have died of cancer. Here again, the Division has developed new epidemiological methods. A third line of research concerns aetiology, aetiogenesis and pathogenesis. Finally, we have a fourth group of projects that do not fit into any of the first three categories. Our research initially focussed on cancer of the urinary bladder and the prostate, and later expanded to include gynaecological cancers. Still, most of our research concerns cancer in the lower pelvic region.

In 2005 Gunnar Steineck initiated the unit at a second location in Gothenburg.

This web site will tell you more about our four lines of research. Summaries of recent dissertations are available, and the researchers who work with us full-time or part-time present themselves and their projects.


Clinical Cancer Epidemiology



Department of Oncology and Pathology
Clinical Cancer Epidemiology
Karolinska University Hospital, Z5:U1 
SE-171 76 Stockholm


Clinical Cancer Epidemiology



Department of Oncology
Division of Clinical Cancer Epidemiology
Onkologiskt centrum
Sahlgrenska University Hospital
SE-413 45 Göteborg 

Long-term follow-up symptom records and quality of life

This part of the program is creating a body of knowledge that may be used to cure patients who have cancer while ensuring that the surrounding organs are not damaged in any absolutely essential way as seen from the patient's point of view. We produce generally applicable scientific knowledge quite independent of the specific technique to be used in future treatment (although we make use of historical material, we are not interested in creating a historical review). Due to a dramatic technical development, we can now raise the level of our ambitions to that of ensuring that all patients successfully treated for cancer can also look forward to a good quality of life in their remaining years.

As concerns radiation treatment the findings presented are intended to show what dosage of ionising radiation delivered to one or more anatomical sites ('organs at risk') provides a specific frequency and intensity for a particular long-term distressing symptom. Within the context of surgery, we address the extent and quality of damage to anatomical sites that lead to specific frequencies and intensities of long-term problems, important for the patient. Progress in molecular biology gives possibility of determining polymorphisms in those genes that control repair in normal tissue after radiation or surgery. The 'hardware' that the technical development gives us must be complemented with 'software' in order for today's clinics to make full use of technical progress, thus ensuring that future patients can be treated successfully. Our program provides this 'software'.

Karin Bergmark

Lars Henningsohn

Massoud al-Abany

Gunnar Steineck

Phone: +46 8 517 75 080 (office)
Phone: +46 8 517 79 621 (fax)



Department of Oncology and Pathology
Division of Clinical Cancer Epidemiology
Z5:U1, Karolinska Hospital
SE - 17176 Stockholm

Phone: +46 31 343 9075 (office)
Phone: +46 31 82 01 14 (fax)



Department of Oncology
Division of Clinical Cancer Epidemiology
Onkologiskt centrum
Sahlgrenska University Hospital
SE - 413 45 Göteborg

Helena Thulin, Nurse

Thomas Hopfgarten

Stefan Buntrock

Gail Dunberger

Helena Lind

Andreas Nilsson

Martin Jonsson

Next of kin

This part of the program identifies avoidable stressors that are unnecessarily induced by health care personnel during a cancer patient's incurable illness and cause psychological traumata and long-term psychological morbidity in the surviving next of kin. Such a stressor can be hiding the next of kin from the fact the patient will die, giving to lack of emotional preparedness, death comes as a shock, leading to long-term anxiety. Another stressor is seeing the patient being insulted by staff. If we succeed in eliminating the identified stressors, the relatives will have a decreased risk of psychological and somatic morbidity, and possibly even mortality, in the long term after a cancer patient's death. For example, if all have an optimal awareness time, or a high degree of emotional preparedness at the time of death, the long-term psychological health among those who lose a wife, a husband, a child, a parent or a sibling due to cancer will improve.

Unnur Valdimarsdóttir

The main focus of my research is on long-term psychological and physical health consequences of stressful exposures, such as the loss of a loved one by death. Of special interest are the modifying effects of forewarning or intellectual and emotional preparedness for such exposures (awareness time).

My PhD work was a population-based study on 379 Swedish widows who had lost their husband to cancer of the prostate or urinary bladder 2-4 years earlier. We found that the widows risk of psychological morbidity at follow-up was mainly explained by two factors additional to the loss itself: the widows short duration of time aware of the impending death (awareness time) and the patients unrelieved psychological symptoms during the last months of life. There is reason to believe that both these factors (e.g. the widows short awareness time and the patients unrelieved symptoms) can be avoided with refined health care.

In co-operation with fellow researchers at Div. Clinical Cancer Epidemiology, I am involved with research on bereaved parents and widowers who have lost due to cancer. Furthermore, we are planning a large study on young adults who have lost a sibling or a parent during childhood or adolescence due to cancer. Together with colleges at the Department of Medical Epidemiology and Biostatistics, I am using the theoretical frame from my previous research to investigate on the impact of loss on development of dementia and the effect of environmental exposures on development of postpartum psychiatric illness. Among others, following hypotheses are being considered:

Widowhood due to cancer

H1: Short time with awareness of husbands impending death from cancer increases the widows risk of physical morbidity, e.g. diabetes, vascular spasm.

H2: A widowers emotional preparedness into the fact of his wifes impending death due to cancer is protective of physical and psychological morbidity in bereavement.

The loss of a child due to cancer

H3: Short time with intellectual and emotional understanding of childs impending death due to cancer increases parents risk of long-term psychological morbidity 4-9 years after loss.

The loss of a sibling or parent due to cancer

H4: The loss of a sibling or parent due to cancer in late childhood or adolescence results in increased risk of mortality owing to both natural and unnatural causes, psychiatric morbidity and lower educational achievement in young adulthood.

Loss and dementia

H5: Sudden unexpected loss of a spouse (as defined in the Swedish Registry of Causes of Death) increases the widow(er)s risk of cognitive impairment/dementia.

Postpartum psychiatric illness

H6: Unexpected developments during pregnancy or labour increase the womans risk for postpartum psychiatric illness.

Ulrika Kreicbergs

Pamela Surkan

Arna Hauksdottir

Ullakarin Nyberg

Aetiology, aetiogenesis and pathogenesis

People exposed to exhaust fumes (from oil, gasoline or diesel fuel) are at increased risk of cancer in the urinary bladder. It is possible that consumption of acetyl salicylic acid (aspirin) protects against this form of cancer. Conversely, consumption of pork and beef probably increases the risk. We examined the hypothesis that heterocyclic amines, compounds that form in the surface when meat is fried or grilled, can cause cancer in humans. These compounds change the genes of bacteria (that is, they are mutagenic) and cause cancer in rats and mice if given at high enough doses. In an epidemiological study we found no evidence that heterocyclic amines cause bladder, kidney, intestinal or rectal cancer. One reason for the lack of evidence of a risk of cancer in humans may be that the daily consumption is so small, on the order of one millionth of the amount given in the animal studies that have shown risks.

The same epidemiological material has been used to study acrylamide, with the equally negative results concerning the risk of cancer in humans, despite acrylamide's ability to change the genes of bacteria (mutagenic capacity). Once more, the reason may be the low doses involved.

Bladder cancer can be viewed clinically (and perhaps also micropathologically) as four different entities in addition to primary cancer in situ. These are: highly differentiated tumours restricted to the urothelium (TaG1, tumours that do not progress within five years of diagnosis); differentiated tumours restricted to the urothelium (TaT2, tumours that almost never prove fatal within five years of diagnosis); poorly differentiated cancer restricted to the urothelium and undifferentiated tumours that invade lamina mucosa (which are fatal in about one-third of the cases within five years of diagnosis, at least with current treatment); and cancer that invades the muscle tissue (for which the mortality rate is probably 100 percent without treatment). The urokinase system is probably important for the development of metastases in bladder cancer. If more than one of the cell cycle proteins are affected, the prognosis is worse than if only one of the proteins ceases to function.

Pancreatitis appears to be caused by pharmaceuticals (for instance drugs used to treat type 2 diabetes) as well as by alcohol consumption and gallstones.

The knowledge about the causes of bladder cancer was a result of Gunnar Steineck's thesis work. Katarina Augustsson-Bälters thesis dealt with whether heterocyclic amines (from fried meat) caused cancer in humans. Maria Seddighzadeh studied urokinase and urokinase receptors. Petra Berggren De Verdier studied the proteins that regulate the cell cycle.

The bladder cancer project is a cooperative venture that involves all of Stockholm's urologists. The publications that have resulted from this joint effort can be seen here.

Somali Sanyal has studed excision repair genes and their relation to the aetiology and course of the disease.

Petra de Verdier

Kerstin Blomgren is studying the causes of pancreatitis.

Kerstin also works at Mälardalens Högskola.

Clinical and methodological studies

A compilation of studies published after 1993 regarding localized prostate cancer: after ten years' follow-up we found that radical prostatectomy reduced mortality to half compared with 'watchful waiting' (i.e. when no therapeutic is action taken immediately after diagnosis of prostate cancer). In absolute terms, the difference was six percent. Radiation therapy, on the other hand, reduced survival compared to 'watchful waiting'. We wondered if the poorer outcome of radiation therapy could be explained by a systematic error in the comparison.

We discussed the issue in a methodological paper that took up 'inverse lead time bias', the lack of comparability that arises when the observation of one group starts two years later (in relation to diagnosis) than the observation of the group they are being compared with. This source of error is conceptually identical to those that can influence a screening study. In another article, we made a precise distinction between factors that predict the outcome of a treatment and factors that determine a prognosis, and exemplified this using data on localized prostate cancer. A prognostic factor is only a factor that predicts treatment effect if the prognostic factor can identify a subgroup of patients who will not have a certain outcome. Follow-up in the Finnish Cancer Registry showed that the outcome death due to prostate cancer essentially disappeared 25 years after diagnosis.

In 1984 Curt Pettersson established a group at Radiumhemmet (Department of Oncology) to study nausea. The group was later broadened to include Sussanne Börjeson and Timo Hursti as doctoral students. Sussanne Börjesons thesis work showed that care-giving during chemotherapy can increase the patients wellbeing, and that it is the duration of nausea, rather than its intensity, that determines to what degree the patient perceives a reduced quality of life. Timo Hursti showed that a subgroup of patients with high endogenous secretion of cortisol had little or no benefit of corticosteroid treatment aimed to reduce nausea caused by chemotherapy. In an extension of this line of research, we are now studying whether acupuncture can reduce nausea caused by radiation therapy. Acupuncture is compared with placebo acupuncture.

Gunnar Steineck has endeavored to summarize and further develop the theoretical basis underlying epidemiological research methods. A study is perceived as having four distinct steps, each of which generates certain systematic errors. Each category of errors affects the relative risks in a way that is determined by rules governing that particular stage. Anna Norder and Susanne Börjeson carry on the studies on acupuncture.

Group members

Gunnar Steineck, Professor

Stockholm office

Lillemor Wallin, Secretary
Else Lundin, ​Research assistant
Erik Onelöv, ​Statistician 
Tommy Nyberg, ​Statistician
Ingela Rådestad, ​Associate professor

Asgeir Helgason, Associate professor
Paul Dickman, ​Associate professor

Ullakarin NybergPost-doc
Unnur ValdimarsdóttirPost-doc
Lars HenningsohnPost-doc
Karin BergmarkPost-doc
Petra de VerdierPost-doc
Massoud al-Abany, Post-doc
Ulrika KreicbergsPost-doc
Anna Bill-Axelsson, Post-doc
Arna Hauksdottir, Post-doc
Anna EnblomPost-doc
Gail DunbergerPost-doc
Helena ThulinPhD student
Thomas HopfgartenPhD student
Stefan Buntrock, PhD student
Helena Lind, PhD student
Eva Johansson, PhD student
Tove Bylund GrenkloPhD student
Andreas Nilsson, PhD student
Pernilla Larsson Omérov, PhD student 
Rossana PettersénPhD student
Martin Jonsson, PhD student

Gothenburg office

Nathalie Ylitalo, Post-doc
Junmei Miao Jonasson, Post-doc
Maria Hedelin, Post-doc
Ann-Charlotte Waldenström, PhD student
Johanna Skoogh, PhD student
Anna Genell, PhD student
Ulrica Wilderäng, Statistician
David Alsadius, PhD student
Dan Lundstedt, PhD student
Þórdís Katrín Þorsteinsdóttir, PhD student
Hanan el Malla, PhD student