Annals of Oncology 12: 1461-1465, 2001. © 2001 Kluwcr Academic Publishers. Printed in the Netherlands.
Original article Quality of life in patients at risk of medullary thyroid carcinoma and followed by a comprehensive medical network: Trends for future evaluations G. Freyer,1 B. Ligneau,1 M. Schlumberger,2 C. Blandy,3 B. Contedevolx,4 V. Trillet-Lenoir,1 G. M. Lenoir,5 N. Chau 6 & A. Dazord 7 'Medical Oncology Unit, Centre Hospitaller Lyon-Sud, and EA 643. Universite Lyon I, Lyon. 2 Department of Nuclear Medicine. Institul Gustave Roussy, Villejuif; 3 Department of Oncogenetics. Institul Curie. Paris; 4 Department of Endocrinology. Hopital de la Timonc, Marseille; 3 Laboratory of Human Genetics. Hopital Edouard Herriot and UMR 5641. Lyon; 6The Lorhandicap Group. INSERM U 420 (National Institute of Health and Medical Research), Faculte de Medecine, Vandoeuvre-les-Nancy, SCR1PT-1NSERM (National Institute of Health and Medical Research), Hopital Saint Jean de Dieu, Lyon. France
Summary Background: As shown in a previous study, the knowledge of the genetic risk in individuals belonging to families at risk of medullary-thyroid carcinoma (MTC) could be associated with impaired quality of life (QoL). Patients and methods: In the present study, we compared the QoL scores obtained in the same period with the subjective quality of life profile (SQLP): in 82 individuals at risk of MTC who had been tested for Ret-mutations; in 200 women at risk of familial breast/ovarian cancer syndrome (BOC); and in a control population of 3501 healthy volunteers. Results: Significant differences were observed in favour of healthy volunteers as well as individuals at risk of MTC, over women at risk of BOC (mean scores: 0.89, 0.85, and 0.64.
Introduction During the last decade we have seen a considerable improvement in our knowledge of familial cancer syndromes. The combination of various malignancies were recognised as 'familial'entities and gene mutations were subsequently identified as being responsible for the occurrence of such syndromes. The availability of specific DNA-mutation tests led to the development of oncogenetic visits which aimed to evaluate the genetic risk of cancer in clinically predisposed families, to offer genetic testing when necessary and to give information to the 'at risk' individuals and to their families about the current recommendations in terms of follow-up and treatment . Medullary thyroid carcinoma (MTC) is familial in around 30% of the cases, due to a germline mutation of the proto-oncogene Ret, located on chromosome 10. This mutation was identified in 1993 as responsible for the occurrence of familial MTC as well as multiple endocrine neoplasia type 2 (MEN 2). Genetic testing is currently available in routine practice in France but the number of families known to be at risk of MTC is small, around 200 to date. All these families are regularly
respectively, P ^ 0.001), but QoL scores were not statistically different between individuals at risk of MTC and the control population (P - 0.2). However, they were significantly inferior in the subgroup of germline Ret-mutation carriers, as compared to the control population (mean scores: 0.73 and 0.89, P = 0.04). In the latter, the relationships with the children and the family were the most important facets of their QoL. Conclusion: Our results confirm the potentially negative impact of the knowledge of the genetic risk of cancer and its consequences in terms of morbidity and follow-up, on the QoL in people followed at oncogenetic visits.
Key words: breast cancer, medullary-thyroid carcinoma, oncogenetics, quality of life, questionnaire
followed by a well-organised medical network including clinicians and geneticists: the Groupe d'Etude des Tumeurs a Calcitonine (GETC). When a Ret-mutation is identified in a given family, genetic testing is offered to people aged five or more. Total thyroidectomy is the reference procedure in germline-mutation carriers. In case of refusal, hormonal follow-up is indicated (repeated thyrocalcitonin assays before and after pentagastrin stimulation). In healthy adults undergoing hormonal follow-up for several years, the fact of being a non-carrier of a germline mutation allows the discontinuation of this procedure . It appears necessary to study the quality of life of people who have undergone genetic testing for MTC: some have been treated for MTC (either sporadic or clearly familial); others are healthy individuals but germline mutation-carriers; others are healthy and noncarriers. Before being informed of the result of the test, it is supposed that all of them are able to transmit any genetic risk to their descendants, and the psychosocial consequences of this uncertainty in intra-familial relationships are not really known . In a previous study we have shown that knowledge of genetic risk could be
1462 associated with poorer quality of life (evaluated with the subjective quality of life profile: SQLP) in germline mutation-carriers, although most of them had already undergone surgery, compared to sporadic MTC patients and healthy non-carriers . In the present study we analysed the same population (/? = 77) plus five new individuals and we compared the quality of life scores with those obtained over the same period using the same questionnaire in a population of women at risk of familial breast/ovarian cancer and with a control population. With this descriptive study our purpose was to provide hypotheses on the relationship between the nature and the potential consequences (follow-up, treatment and risk of mortality) of a genetic risk, and an individual's quality of life.
Patients and methods Population characteristics Individuals at risk of MTC Eighty-two consecutive subjects entered the study. They were followed in the Department of Nuclear Medicine, Institut Gustave Roussy. Villejuif (n = 73) or in the Endocrinology Department. Hopital de la Timone, Marseille (n - 9). two centres belonging to the GETC. All had been followed and/or treated for sporadic or familial MTC, or should belong to a predisposed family. Exclusion criteria were: evolutive MTC, recent ( < 3 months) surgery, chemotherapy, radiotherapy or brachytherapy. inability to fill in a questionnaire. Individual covariates which could influence the quality of life scores were recorded, presence or absence of a Ret germline-mutation; hormonal follow-up or not: hislological findings when available (no disease. C-cell hyperplasia or cancer), time elapsed since the beginning of the follow-up and since surgery for patients who had undergone total thyroidectomy; other signs of MEN type 2A (pheochromocytoma. hyperparathyroidism) or 2B (marfanoid habitus, ganglioncuromatosis, mucosal neuromas). Among the 82 subjects who entered the study, the youngest was 11 and the oldest 77 years old (mean = 46.3). There were 36 men and 46 women. Surgery had been performed in 65 patients: 60 were found to have MTC and live had C-cell hyperplasia only. No patient had any cancer-related symptom at the time of the study. Forty-five patients were sporadic cases and 37 subjects were MTC-family members. In this last population. 20 subjects were germline- mutation carriers and had already undergone total thyroidectomy for MTC (n = 15) or C-cell hyperplasia (n - 5), 15 were germline mutation carriers but still did not undergo surgery. In fact, these 15 subjects belong to a large family with a mutation at codon 804. which is known to be associated with a late onset of the disease and with low aggressiveness. Two participants were non-carriers and healthy individuals. There was only one case of MEN 2A with a surgically-removed pheochromocytoma. but no case ofMEN2B.
treatment at the time the study began. In 29%. the treatment was completed less than two years before study entry , Control population (healthy individuals) This sample consisted of all the people aged 15 or more from 8.000 households (entity defined by a global consumption: family or person living alone) randomly selected from the Lorraine area (north eastern France: 2 3 million inhabitants). Over three months before the survey, a campaign was conducted through the media (television, newspapers, radio) to inform the population of the survey. The selection of the households was made from those with a telephone. The protocol of the study included: 1) the completion of a participation request form including a questionnaire to determine the number of people in the household. 2) the mailing of a standardised auto-questionnaire (the SQLP described below) with a cover letter and a pre-paid envelope for its return, 3) a simple follow-up letter; 4) further follow-up with questionnaires, a cover letter and a pre-paid envelope. These documents were addressed to the households with a letter explaining the objective of the research at one month intervals. When the number of individuals in a given household was unknown, more questionnaires were sent when necessary, according to the answer to the first one. Among the 8000 households selected. 6269 individuals finally replied. We selected 3501 subjects who considered their own health status as 'good' and explicitly indicated that they had never been treated for cancer. These subjects formed our "control population". Their mean age was 39 (15-80); 49.3% were men. 50.7% women.
The Subjective Quality of Life Profile (SQLP) This questionnaire was developed by SCRIPT-1NSERM (National Institute of Health and Medical Research. Lyon. France) and both the French and English versions were subsequently validated [6. 7]. It may be used for the assessment of quality of life, both in healthy individuals and in patients, whatever their disease. It allows investigators to explore the various facets of the quality of life of the patients. It is divided into a 'core-questionnaire" which must be entirely filled in by each patient and an "optional questionnaire' which includes a variable number of items chosen by the investigator from a 'bank', depending on the particular points he wishes to focus on (e.g. family, children etc.). The use of the core-questionnaire is necessary for the overall statistical validity of the analysis. For each item of the SQLP. the participant is invited to express his/her degree of satisfaction using a score varying from - 2 (very dissatisfied) to +2 (very satisfied). Intermediate scores are - 1 : somewhat dissatisfied. 0: indifferent. +1: somewhat satisfied This question design enables the investigator to evaluate various aspects of the quality of life such as physical and psychological well-being, inter-individual and social relationships. These aspects are presented as scores, the total score being the mean of all sub-scores. This is a very simple and useful presentation that allows the usual statistical tests and comparisons between sub-populations to be performed. Moreover, the influence of discrete as well as continuous covariates such as age, sex. pathology... on the quality of life scores can readily be studied , Table 1 summarises the items used in the present study
Individuals at risk for hreasl/ovarian carcinoma Statistical analysis Two hundred women at risk for familial breast/ovarian carcinoma referred to the oncogenetic unit of the Institut Curie, Paris, participated in the study. Eligibility criteria were: healthy subject or cancer patient belonging to a family at risk for breast/ovarian carcinoma (at least three cases in relatives of the first degree), or occurrence of breast cancer before the age of 36 in the proposant, with or without familial history. Exclusion criteria were: inability to fill in a questionnaire, life expectancy < 3 months. The mean age of the population was 44 years (25-65). One hundred and thirty one women had breast cancer and 10 had ovarian cancer. The others were healthy Among cancer patients, 26% were still under
The means of the various sub-scores as well as the total scores were calculated for each population (individuals at risk of MTC. women at risk of familial breast/ovarian carcinoma and the control population) and compared by using the Student's /-test. Within the population of individuals at risk of MTC, some subgroups were defined according to the following covariates: presence vs. absence of the inherited predisposition, surgery vs. no surgery performed. The 'cancer' vs. 'no cancer' covariate was used in the population of women at risk for familial breast/ovarian carcinoma. Sex and age were used as covariates in all three populations. Comparisons were performed between all the
1463 Table 1. The subjective quality of life profile (SQLP)- list of the 30 items used in the study. I. 2 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.
Your physical abilities Your intellectual abilities Your autonomy Your overall fitness Your physical appearance Your sexuality Your education The fact that you have (or do not have) a professional activity The assistance of your family Your independence from your family Your relationships with your family Your relationships with your colleagues at work Your relationships with your friends People's attitude towards you Your meals, what you can eat Your income or the money that you have available Your opinion about yourself The fact that you have (or do not have) children The education of your children The fact that you participate (or do not participate) in group activities The fact that you live (or do not live) with someone (as a couple, as a family, as a single person) The fact that you have someone you can trust Your personal or inner life (meditation, daydreaming, reading, reflection...) The fact that you have (or do not have) a religious faith The fact that you feel (or do not feel) free The place where you live Your personal situation The quality of your daily life The way you see your future What is happening in the world
subgroups. The results were considered as statistically significant for P ^ 0.05.
Scores by items Table 2 shows the statistically significant differences observed in various items between the germline mutation-carriers, the non-carriers and the control population. Most significant items concerned either the familial environment (relationships with the family, independence, children) or the individual well-being (physical and intellectual abilities, overall fitness). Individuals at risk for familial MTC vs. breast-ovarian cancer (BOC) Overall scores The results for both the individuals at risk for MTC and the control population are presented above. The mean score in the individuals at risk for BOC was 0.64. The differences between this score and both that of people at risk of MTC (P - 0.001) and of the control population (P < 0.001) were significant. When the presence of a germline mutation and the individual antecedents of breast/ovarian carcinoma (BOC patients) were taken into account, the following results were obtained: Germline Ret-mutation carriers: 0.73; BOC patients: 0.60; Healthy women at risk of BOC: 0.71; All the differences between these subgroups were significant (P < 0.05) except that between BOC patients and women at risk of BOC (P = 0.07). Influence of individual covariates Neither age nor gender had evident impact on the observed differences. Scores by items Table 3 shows the significant differences observed in various items between the women at risk of BOC and the people at risk of MTC. There were seven significant differences, all in favour of the latter.
Individuals at risk of familial MTC vs. control Discussion Overall scores There was no statistically significant difference between the overall mean scores in the population of individuals at risk of familial MTC (n — 82) and in the control population (/? = 3501): 0.85 vs. 0.89, respectively (P = 0.2). In contrast we observed a significant difference between the germline mutation-carriers and the control population: the mean scores were 0.73 and 0.89, respectively (P - 0.04). Influence of individual covariates As previously reported , the fact of having been surgically treated for MTC or C-cell hyperplasia had no impact on the quality of life scores. Interestingly enough, there was a significant difference (P = 0.001) between men (mean score = 0.97) and women (mean score = 0.69).
There are several factors that could at least partly influence the quality of life in people involved in familial cancer syndromes. First, medical knowledge about the mode of inheritance, the corresponding DNA-mutations that can readily be explored in most cases, the level of risk for the individuals and the consequences in terms of follow-up and surgical prophylactic procedures has markedly increased in the last decade. Thus, the information given to people attending genetic counselling has become much more precise and is likely to induce more stress in some cases. Second, the aggressiveness of some 'prophylactic procedures', such as bilateral mastectomy and oophorectomy in women at risk of breast/ovarian cancer syndrome is associated with a substantial alteration of their quality of life . The follow-up itself may be constraining: repeated colonoscopies in subjects at risk of hereditary non-polyposis colorectal carcinoma (HNPCC), without certainty about curability when can-
1464 Table 2 Differences in subscores between individuals at risk for MTC and the control population. Item
Germline mutation carriers (mean score)
Non carriers (mean score)
(Control population I mean score)
1. 2. 4. 10. II. 13. 14. 15 18.
0.47 (S) 0.66 (S) 0.56(S) 1.32 (S) 0.53 (S) 0.71 (S) 0.84 (NS) 1.03 (NS) 1.48 (S)
0.43 (S) 0.65 (S) 0 66(S) 1.23 (NS) 1.02 (NS) 1.20 (NS) 1.14 (S) 1.28 (S) 1 43(S)
.03 .02 23 .03 .22 .22 ()85 04 .03
< 0.0001 0.008 < 0.0001 0.048 < 0.0001 < 0.0001 0.013 0.049 0.015
Physical abilities Intellectual Abilities Overall Fitness Independance/Family Relationships/Family Relationships/Friends People's attitude/you Your meals Having children (or not)
The number of each item refers to Table 1. S - significant (P < 0.05): NS - non-significant (P > 0.05). All the levels of significance are presented by reference to the control population. A'-value significant difference; when there are two significant differences, the highest value o f f is indicated.
Table 3 Significant differences in subscores between individuals at risk for medullary-thyroid carcinoma (MTC) and women at risk for breast-ovarian carcinoma (BOC) Item
MTC: mean scores
1. 4. 6. 10. 20. 23. 28.
0.44 0.64 0 75 1.28 0.55 0 88 1.06
Physical abilities Overall fitness Your sexuality Independance from family Group activities Inner life Quality of daily life
BOC: mean scores -0.07 0.23 0.39 0.90 0.28 041 0.41
0.001 0 007 0.026 0.001 0.039 0.001 < 0.001
cer occurs, gives us a good example . Third, the delay between blood sampling and the results of genetic testing is often long enough to let doubt and anxiety develop in many people. Fourth, the fact of being the 'messenger of the (bad) news' for a patient, e.g. the necessity for him to inform all the family members that they are potential 'risk-carriers' is certainly deleterious. For all these reasons, a number of authors have underlined the importance of quality of life studies performed in people at risk of familial cancer, or from a more general point of view, in people attending oncogenetic visits . In this perspective, MTC is an interesting tumour model since genetic testing is simple and reproducible, hormonal follow-up is efficient and readily accepted by the patients, and prophylactic surgery is associated with low morbidity and a high long-term cure rate. We have previously reported that the fact of being a germline Ret-mutation carrier could be associated with a substantial decrease in quality of life, with respect to other major prognostic factors such as occurrence of MTC (vs. no cancer) and total thyroidectomy (vs. no surgery performed). Our results suggested that it could be due to the proposant's knowledge about the disease and to its consequences on the intra-familial relationships . From this point of view, the 'curative' effect of prophylactic thyroidectomy does not overcome those difficulties. The methodological difficulties of such a study include
the heterogeneous nature of the population. In this study, the mean age of the three populations was different, individuals from the control population being the youngest. Similarly, there were 57% women in the MTC group, 100% in the BOC group and 50.7% in the control population. Thist could bias the results of the analysis. To overcome this limitation, we used a measurement tool (the SQLP) that allows individual covariates to be taken into account. Neither age nor gender seemed to have an influence on the comparisons performed between the three populations. Moreover, the SQLP does not focus on particular problems of cancer patients, such as treatment toxicity or cancer-related pain; it is well adapted to the quality of life assessment, both in healthy individuals and patients, whatever their disease. In the present study, we wished to compare two different populations at risk for familial cancer followed by oncologists and geneticists involved in oncogenetic visits. We assumed that the known differences in the above-mentioned quality of life 'prognostic factors' between those two populations - women at risk for BOC are supposed to be much more distressed - could result in measurable differences in quality of life scores. Assuming the heterogeneity of these populations and the peculiarities of each tumour type, this comparison is not inferential and simply provides a basis for further studies. That is why all statistical comparisons should be performed by reference to a control population of healthy individuals who were studied over the same period. The comparison between Ret-mutation carriers, non-carriers and healthy individuals from the control population confirms, in a larger population of study participants, our previous findings concerning the deleterious impact of the knowledge of the genetic risk. In the precedent study, we had performed a direct comparison between carriers and non-carriers; in the present study, this comparison is indirectly validated by reference to the control population. It is noteworthy that we found no significant difference between control individuals and non-carriers, despite the fact that most of them had been treated for MTC. As far as the subscores are concerned, most of the
1465 differences involve health, family and relationships within the social group (friends, other people). Ret-mutation carriers express more concern about their independence from the family and about children. These particular aspects of their life appear more crucial than to other groups. The comparison between subjects at risk for MTC and BOC should be interpreted with caution, since some of the women at risk for BOC had been treated for cancer within the last two years, or were still under treatment. However, the difference between those patients and the other women (at risk but healthy) was not significant. From a general point of view, among people attending oncogenetic visits, better quality of life scores, as measured by SQLP, were obtained in people at risk of MTC rather than BOC. The differences are much more obvious in seven of the subscores, all in favour of individuals at risk for MTC. We must stress that the way information about predisposition to cancer is usually given does not substantially differ between the two groups of investigators involved in this study (the GETC and the Institut Curie); individual follow-up and psychosocial support are also similar in both groups. However, there is no major difference between the Ret-mutation carriers and the healthy women at risk for BOC, whose mean QoL scores are low ( < 0.75). This could suggest, one more time, the negative impact of knowledge of a familial risk of cancer in people who are preoccupied not only by their own health, but also by their relationship with their parents and the consequences of their condition for their offspring [12-14].
B. Legras, A. Caria, A. Dazord, L. Mejean, N. TubianaRufi, J. P. Meyer, J. C. Cnockaert, M. V. Ticheur, G. Bourgey, P. Redor, M. D. Wagnon,Y. Schleret, J. M. Mur. The group is animated by N. Chau. They also wish to thank Ms Alison Foote for careful reading of the manuscript.
References 1 2
Conclusion The results of our study confirm the negative impact of the knowledge of the genetic risk of cancer on the quality of life in some cases. They also suggest the existence of a relationship between quality of life impairment and knowledge of the consequences of a predisposition to cancer in terms of follow-up, prophylactic procedures, and treatment. However, our study was not inferential and thus did not clearly demonstrate such a relationship. This particular point needs to be addressed in future studies. Psychosocial follow-up and support should be offered to all people at risk of familial cancer and undergoing genetic testing.
Collins FS. Ethical, legal and social implications of DNA testing for cancer predisposition. Anti-Cancer Res 1994: 10: 210-5. SchufTenecker I. Billaud M. Calender A. RET proto-oncogene mutations in French MEN 2A and FMTC families. Hum Mol Genet 1994: 3: 1939-43 Grosfeld FJM, Lips CJM. Beemer FA. Psychosocial consequences of DNA analysis for MEN type 2. Oncology 1996: 10 141-6. Freyer G. Dazord A. Schlumberger M, Contedevolx B ct al. Psychosocial impact of genetic testing in familial medullarythyroid carcinoma: A multicentric pilot-evaluation. Ann Oncol 1999; 10: 87-95 Blandy C. Schwab S. Stoppa-Lyonnet D. Duzord A. Psychosocial impact of Ihe first consultation in oncogenetics on women with a familial history of breast and ovarian cancer. Bull Cancer 1998: 7: 637-43. Gerin P. Dazord A. Boissel JP. Chifilet R. Quality of life assessment in therapeutic trials Rationale for and presentation of a more appropriate instrument. Fundamental and clinical Pharmacology 1992; 6. 263-76. Bonami A. Patrick DL. Bushnell DM. Martin M. Validation of the united States' version of the World Health Organization Quality of Life (WHOQOL) instrument. J Clin Epidemiol 2000; 53: 1-12. Dazord A. Quality of life assessment in Medicine. Presentation of a new instrument (SQLP. Subjective quality of life profile) and first results. Eur Respir Rev 1995: 5- 66-71. King MC. Rowell S. Love S. Inherited breast and ovarian cancer. What are the risks? What are the choices? JAMA 1993: 15: 197580. Marra G. Boland CR Hereditary non polyposis colorectal cancer: The syndrome, the genes, and historical perspectives. J Nail Cancer Inst 1995:87: 1114-25. Michie S. Mac Donald V. MarteauT Understanding responses to predictive genetic testing A grounded theory approach. Psychol Health 1997; 14: 327-34. Lynch H. Lynch J. ConwayT. Severin M. Psychological aspects of monitoring high risk women for breast cancer. Cancer 1994: 74' 1184-92. Redelmeier DA. Rozin P. Kahneman D. Understanding patient's decisions. Cognitive and emotional perspectives. JAMA 1993; 270: 72-6. Voelker R. The genetic revolution- Despite perfection of elegant techniques, ethical answers are still elusive. JAMA 1993; 270: 2273-4.
Received 29 October 2000; accepted 8 May 2001.
Acknowledgements The authors wish to thank the Lorhandicap Group which conducted the study in the Lorraine population. Its members are F. Guillemin, J. M. Andre, J. F. Ravaud, J. Sanchez, S. Guillaume, J. P. Michaely, C. Otero-Sierra,
Correspondence lo G. Freyer, MD, PhD Medical Oncology Unit Centre Hospitalier Lyon-Sud 69495 Pierre-Benite cedex France E-mail. Gilles.Freyer(«)chu-lyon.fr