U.S. flag

An official website of the United States government

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Diagnosis and Management of Metastatic Malignant Disease of Unknown Primary Origin. Cardiff (UK): National Collaborating Centre for Cancer (UK); 2010 Jul. (NICE Clinical Guidelines, No. 104.)

Cover of Diagnosis and Management of Metastatic Malignant Disease of Unknown Primary Origin

Diagnosis and Management of Metastatic Malignant Disease of Unknown Primary Origin.

Show details

Guideline chapter 3Factors Influencing Management

13. Investigations to find the primary tumour in people with cancer of unknown primary, when clinical benefit is unlikely

Last updated: 30/10/2009.

Short summary

There is evidence that people with CUP sometimes receive excessive diagnostic evaluation (Shaw et al, 2007). Diagnostic investigations limited to fewer tests would not affect survival in most patients, but this could have a negative impact on patients’ psychological well being.

Very few studies reported the psychological effect of diagnosis of the primary tumour in people with CUP. The best evidence came from a qualitative study of a small group of people with CUP (Boyland and Davis, 2008). There was evidence that people with cancer of unknown primary experience uncertainty and distress. Patients have to deal with the uncertainty about the origin of their disease, its future course and the benefit of treatment.

In most cases finding a primary is unlikely to significantly improve outcome, but this appears contrary to patients’ beliefs. Some patients felt that they were missing the chance of targeted therapy if their primary is not found. Patients with at least a suspected primary site gained some benefit in being able to focus on their treatment plan.

No studies directly compared minimal versus exhaustive diagnostic evaluation in terms of patients’ quality of life.

Rationale

Conventional medical management of patients with malignancy of undefined primary origin concentrates on undertaking a minimum set of investigations to try and define a primary tumour site, with a view to providing rationally based treatment. A specific aim is to avoid “futile” or protracted investigations when the likelihood of further clarifying the diagnosis has become very low. This approach neglects an important priority for some patients, which is to gain the highest possible certainty about the nature of their illness, regardless of the extent of investigations which have to be performed.

In some instances, an explanation of the strategy, and the limitations of further tests will satisfactorily allay a patient’s concerns. In other cases there may be remaining uncertainty, causing psychological morbidity, which in the patient’s mind can only adequately be addressed by further tests seeking a possible primary, regardless of the low yield and additional inconvenience. To optimise the care of patients with malignancy of undefined primary origin it is necessary to try and define the optimal point for ceasing diagnostic tests, based on a balance between standard clinical benefit and individual psychological need.

Methods

Study types

There was no restriction on study design.

Participants

People with malignancy of undefined primary origin in the initial diagnostic phase and people with confirmed cancer of unknown primary origin at the completion of standard investigations.

Interventions

Further investigations to try and find the primary, compared with no further diagnostic tests.

Outcomes

Patient’s psychological adjustment. Clinicians confidence in their ability.

Study selection

An initial list of studies was selected by the information specialist (SA). One reviewer (NB) then selected potentially relevant papers from this list on the basis of their title and abstract. These studies were ordered and each paper was check against the inclusion criteria.

Data extraction and synthesis

One reviewer (NB) extracted data. Qualitative data was summarised by listing the themes identified in the studies. Patient’s first hand experiences about uncertainty and the diagnostic process were also included when available.

Search results

The literature search identified 14 studies, six of which were included. An additional study (Shaw et al, 2007) was included as evidence of the typical diagnostic evaluation of people with CUP in the UK.

Description of included studies

The studies included a qualitative study of ten patients with CUP (Boyland and Davies, 2008), a study of psychological adjustment in a group of 72 patients with CUP (Lenzi et al, 2004) and three expert reviews (Chorost el al, 2004; Ettinger 2005; Symons, 2008).

Evidence summary

Shaw et al (2007) reviewed the investigation and management of carcinoma of unknown primary in a single UK cancer network during 2003. A wide variety of tests were used in the diagnostic evaluation of these patients, either before or after referral to the cancer centre. Nineteen different investigations were used in the cohort of patients with liver or multiple metastases, 13 different tests were used in the cohort with bone metastases. Shaw et al (2007) concluded that the number of diagnostic investigations could be reduced substantially, suggesting tests should be limited to those affecting clinical management.

Qualitative evidence (Boyland and Davis, 2008)

Boyland and Davis (2008) identified six main themes in their study: poor understanding, struggling with uncertainty (contrasting with stoical acceptance), undergoing multiple investigations, inability to treat, healthcare professionals not having the answers and difficulty explaining to others.

Understanding of CUP

All patients with an entirely unknown primary reported being told they had cancer but that the primary site could not be found. Some patients did not fully understand this.

“This kind of non-specific kind of ... they haven’t found the primary tumour but it is spreading all over the place.”

Uncertainty about diagnosis

Patients clearly struggled with the unknown nature of the primary tumour. This seemed to increase the unpredictability of the disease, with patients not knowing what to expect, feeling an ominous sense that it might be “spreading” or “lurking”.

“I think that if there is a secondary and it can cause you that much jip, if there is is a primary it could do you double the damage. I just don’t understand how it can hide away somewhere...it’s the not knowing is the horrible thing ...the uncertainty of it all...[if I knew] I would be more at ease.”

Others wanted the understanding and feeling of control attached to a diagnostic label.

“Its confusion because you don’t know what to expect. I know there are loads of cancers around and they know where most of them are, well why I am so different? Why are these unknown primaries? So ... I feel like screaming, literally screaming. [If] they said where they are ... well, for me it would be peace of mind.”

One man had been diagnosed with leukaemia 20 years earlier and was able to compare the experiences of having known and unknown cancer.

“I’ve got no feeling where the actual cancer is and (my wife) quite often has a prod to see if she can find it. With the cancer I had before I knew exactly where it came from, but not knowing with this cancer makes me like unaware and I would like to know where it has come from.”

Some patients accepted that their primary was unknown and that there was no point in thinking about finding it.

“... if it is there it is there. I mean it doesn’t make any difference to me no ... so trying to think about it is to me a bit of a waste of time.”

One patient, with a possible ovarian primary, found it useful to believe it was an ovarian primary.

“As far as I’m concerned it is in my ovaries ... because I’m being treated for ovarian cancer. I’m not looking for anything else at the moment. It would be much more difficult if I didn’t know where it was.”

Multiplie diagnostic tests

All participants experienced a series of unsuccessful tests to find the primary tumour:

“...a whole series of tests, CT scans, MRI - you name it I had it...and in the end they said well, we can’t trace it.”

“They seem to have covered the whole of my body with tests and things.”

Finding the primary and targeting treatment

Many patients believed that finding the primary tumour would lead to more effective treatment.

“If they knew where it was they’d be doing something about it. I mean they have told me that they cannot do anything about it at all, it’s only palliative and I can accept that.”

Several patients felt that they were receiving untried and untested treatment.

“She said... they have not done that mixture before, so the side effects might cancel each other out or make it worse ... not sure about long-term effects ... very high dosage.”

In contrast the patient with suspected ovarian cancer was more reassured by her treatment plan.

“They said we’re going to treat you for ovarian cancer as that is the direction the tests are pointing, so as far as I was concerned that was it, a plan was in place. Because I’ve got a plan I’m concentrating on that, not on the negative.”

The uncertainty of healthcare professionals

All patients referred to the uncertainty of the healthcare professionals involved in their care.

“I do understand they are in the dark as much as me...They don’t know enough about this unknown primary situation. Perhaps that’s why they don’t tell you much because they are not sure of what they are telling you.”

One patient was worried that the consultants were “baffled”, but another acknowledged the difficulty faced by healthcare professionals in the diagnosis and treatment of patients with CUP.

Psychological adjustment

In their study, Lenzi et al (2004) reported that people with CUP had higher levels of uncertainty than other patients with cancer, but did not present supporting data. They also reported over 40% of patients showed signs of depression. Other expert reviews (Symons, 2008; Ettinger, 2005) suggest that increasing patient’s knowledge about their diagnosis can help dispel some of these fears.

In their questionnaire study, Pirian et al (2005) asked 45 American patients to imagine they had metastatic cancer of unknown origin. Patients were willing to pay a average of $1900 for ancillary immunohistochemical tests to identify a primary tumour, even when these tests would not affect their survival.

References
  • Boyland L, Davis C. Patients’ experiences of carcinoma of unknown primary site: dealing with uncertainty. Palliative Medicine. 2008;22(2):177–83. [PubMed: 18372382]
  • Chorost MI. Unknown primary. Journal of Surgical Oncology. 2004;87(4):191–203. [PubMed: 15334635]
  • Ettinger DS. Occult primary cancer: Clinical practice guidelines. JNCCN Journal of the National Comprehensive Cancer Network. 2005;3(2):214–33. [PubMed: 19817031]
  • Lenzi R, Abbruzzese JL, Baile WF, Cohen L, Parker PA. A study of psychological adjustment in patients with metastatic cancer of unknown primary. Psycho-Oncology. 2004;13(8 Suppl):357.
  • Pirain DM, Gryzbicki DM, Andrew-Ja-Ja C, Raab SS. Measuring patient preferences for ancillary testing: patient willingness-to-pay for immunohistochemistry in tumors of unknown primary. Modern Pathology. 2005;18(Suppl 1):324A–325A.
  • Shaw PHS, Adams R, Jordan C, Crosby TDL. A clinical review of the investigation and management of carcinoma of unknown primary in a single cancer network. Clincial Oncology. 2007;19:87–95. [PubMed: 17305260]
  • Symons J. Supporting patients with cancer of unknown primary. Nursing Times. 2008;104(14):23–4. [PubMed: 18497236]

Characteristics of included studies

Boyland-2008

Chorost-2004

Ettinger-2005

Lenzi-2004

Pirain-2005

Shaw-2007

Symons-2008

References for included studies

Boyland 2008.
Boyland L, Davis C. Patients’ experiences of carcinoma of unknown primary site: dealing with uncertainty. Palliative Medicine. 2008;22(2):177–83. [PubMed: 18372382]
Chorost 2004.
Chorost MI. Unknown primary. Journal of Surgical Oncology. 2004;87(4):191–203. [PubMed: 15334635]
Ettinger 2005.
Ettinger DS. Occult primary cancer: Clinical practice guidelines. JNCCN Journal of the National Comprehensive Cancer Network. 2005;3(2):214–33. [PubMed: 19817031]
Lenzi 2004.
Lenzi R, Abbruzzese JL, Baile WF, Cohen L, Parker PA. A study of psychological adjustment in patients with metastatic cancer of unknown primary. Psycho-Oncology. 2004;13(8 Suppl):357.
Pirain 2005.
Pirain DM, Gryzbicki DM, Andrew-Ja-Ja C, Raab SS. Measuring patient preferences for ancillary testing: patient willingness-to-pay for immunohistochemistry in tumors of unknown primary. Modern Pathology. 2005;18(Suppl 1):324A–325A.
Shaw 2007.
Shaw PHS, Adams R, Jordan C, Crosby TDL. A clinical review of the investigation and management of carcinoma of unknown primary in a single cancer network. Clincial Oncology. 2007;19:87–95. [PubMed: 17305260]
Symons 2008.
Symons J. Supporting patients with cancer of unknown primary. Nursing Times. 2008;104(14):23–4. [PubMed: 18497236]

14. Prognostic and predictive factors in CUP

Last updated: 30 /10 / 2009.

Short summary

There was evidence that certain factors are associated with response to chemotherapy and overall survival in people with CUP.

While many prognostic factors appeared important on univariate analysis, few remained so on multivariate analysis. Independent adverse prognostic factors included: presence of liver metastases, low serum albumin and elevated serum lactate dehydrogenase. Good performance status was the only independent favourable prognostic factor consistently reported in studies.

Several authors have developed simple prognostic models incorporating some of these factors to which can classify people with CUP into low and high risk groups. These risk groups have statistically significant differences in overall survival, but their clinical significance is unclear: there are no studies evaluating whether these prognostic models influence treatment decisions. There is inconsistency between the factors included in the prognostic models, suggesting differences between the populations used to develop them

There was a lack of prognostic models to estimate the absolute survival probability of a given patient with CUP.

Rationale

For all cancer patients, the decision to introduce treatment is based on the balance of costs (toxicity, inconvenience) and benefits (relief of symptoms, prolongation of survival). The same principle applies to confirmed Cancer of Unknown Primary, though the more limited efficacy of treatment means that the greatest care should be taken in weighing the factors in these patients. In confirmed CUP, accurate prognostic predictors are potentially of great value in clinical decision making, allowing optimal treatment to be used in those most likely to gain the greatest benefit, while avoiding the unnecessary toxicity of futile treatment in those unlikely to benefit.

Individual physiological factors influence the likelihood that an individual will tolerate chemotherapy toxicity, and to a certain degree also influence the likelihood of benefit. These factors include organ function, performance status and co-morbidity. Tumour-specific factors (e.g. chemosensitivity, tumour burden, specific organ involvement) partly govern the likelihood of a satisfactory outcome of treatment. In many instances the factors referred to are unknown, or difficult to measure.

Defining major prognostic factors governing treatment outcomes in confirmed CUP would be of considerable benefit, both in terms of individual patient care, and more widely in terms of avoiding unnecessary treatment costs where such treatment could be predicted to be futile.

Methods

Study types

Any studies reporting prognostic analysis in patients with CUP, there was no restriction on study design.

Participants

People with confirmed Cancer of Unknown Primary in whom systemic therapy is being considered. Studies restricted to patients with a single specific presentation (such as squamous cell carcinoma in cervical lymph nodes) were excluded.

Interventions

Prognostic factors with an established role in general cancer treatment including: performance status, age, LDH, tumour burden and critical organ involvement.

Outcomes

Treatment outcomes: overall survival, treatment response. Change in management and avoidance of inappropriate treatment

Study selection

An initial list of studies was selected by the information specialist (SA). One reviewer (NB) then selected potentially relevant papers from this list on the basis of their title and abstract. These studies were ordered and the reviewer checked each paper against the inclusion criteria. Studies ordered for previous questions about chemotherapy were also checked for prognostic factor analysis.

Data extraction and synthesis

One reviewer (NB) extracted outcome data from the papers. Treatment response was treated as a dichotomous variable: any response or no response, and summarised using risk ratios. Overall survival data include both the event (death from any cause) and the time at which the event occurs. Time-to-event outcomes are most appropriately analysed using hazard-ratios (HRs) which incorporate both the number and the timing of events. Overall survival was analysed using methods outlined in Tierney et al (2007). In most cases the log-rank P value and the overall death rate were the only data available to estimate the hazard ratio. The data from each study were pooled using the generic inverse variance method in the Cochrane RevMan software package.

Quality assessment

Study quality (risk of bias) was assessed using the NICE checklists for critical appraisal.

Heterogeneity assessment

Heterogeniety was assessed in Forest plots using the I-squared statistic.

Search results

The literature searches identified 103 potentially relevant studies, of which 50 were included.

Description of included studies

Six studies reported prognostic factor analyses in patients with CUP, regardless of their treatment (Abbruzzese et al, 1994; Hess et al, 1999; Ponce Lorenzo et al, 2007; Seve et al, 2006; Trivanovic et al, 2009; Van de Wouw et al 2004). The remaining papers described case series or clinical trials in which the majority of patients received chemotherapy. Data about predictive factors for treatment response were drawn from these chemotherapy studies.

The studies typically excluded patients with cancer of unknown primary belonging to a subgroup with well defined treatment. In most cases histology was well, moderately well or poorly differentiated adenocarcinoma or poorly differentiated carcinoma. Abruzzesse et al (1994), Hess et al (1999), Jentsh-Ullrich et al (1998) and Hainsworth (1997), however, included patients with other histology. Van der Gast et al (1995) included only patients with undifferentiated carcinoma or poorly differentiated adenocarcinoma of unknown primary.

Most studies reported at univariate analysis of prognostic factors and many also reported multivariate analysis. Univariate analyses consider a single prognostic factor at a time, often splitting the patient group into two and comparing the outcomes of patients with and without the factor. Prognostic factors are not necessarily independent, for example elevated serum alkaline phosphatase and bone metastases are probably correlated. For this reason relative risks associated with multiple individual prognostic factors from univariate analyses cannot be combined to give an overall risk score.

Multivariate analysis is more useful as it estimates the independent effect of each factor. Thus several prognostic factors can be combined to estimate the absolute risk or probability than an event will occur in a given patient. Multivariate analysis form the basis for the prognostic models developed in some of the studies (see table 14.3).

Table 14.3. Predictive factors for treatment response, risk ratio and 95% confidence interval.

Table 14.3

Predictive factors for treatment response, risk ratio and 95% confidence interval.

Some chemotherapy trials reported individual patient data for those who responded to chemotherapy, allowing univariate analysis of predictors of treatment response.

Study quality

Some studies using multivariate analyses only reported prognostic factors that were statistically significant. This reporting bias could lead to an overestimation of the effect of a given prognostic factor when pooling the results of these studies.

Continuous or ordinal prognostic variables (such as age, LDH level, performance status or number of involved sites) were typically dichotomised into high or low groups using an arbitrary cut-point. This could underestimate the effect of these prognostic factors. The location of the cut point can also be influenced post-hoc by the data, by choosing a cut point which maximises the effect of the prognostic factor.

Evidence summary

Prognostic factors

Prognostic factors for overall survival and predictive factors for treatment response are summarised in tables 14.1 to 14.3, and in figures 14.1 to 14.31.

Table 14.1. Prognostic factors investigated using multivariate analysis.

Table 14.1

Prognostic factors investigated using multivariate analysis.

Table 14.2. Prognostic factors for overall survival, hazard ratio and 95% confidence interval.

Table 14.2

Prognostic factors for overall survival, hazard ratio and 95% confidence interval.

Figure 14.1. Well or moderately well differentiated adenocarcinoma versus other histology, univariate analysis of treatment response.

Figure 14.1

Well or moderately well differentiated adenocarcinoma versus other histology, univariate analysis of treatment response.

Figure 14.2. Well or moderately well differentiated adenocarcinoma versus other histology, univariate analysis of overall survival.

Figure 14.2

Well or moderately well differentiated adenocarcinoma versus other histology, univariate analysis of overall survival.

Figure 14.3. Poorly differentiated adenocarcinoma or carcinoma versus other histology, univariate analysis of treatment response.

Figure 14.3

Poorly differentiated adenocarcinoma or carcinoma versus other histology, univariate analysis of treatment response.

Figure 14.4. Poorly differentiated adenocarcinoma or carcinoma versus other histology, univariate analysis of overall survival.

Figure 14.4

Poorly differentiated adenocarcinoma or carcinoma versus other histology, univariate analysis of overall survival.

Figure 14.5. Undifferentiated carcinoma, univariate analysis of treatment response.

Figure 14.5

Undifferentiated carcinoma, univariate analysis of treatment response.

Figure 14.6. Male versus female, univariate analysis of treatment response.

Figure 14.6

Male versus female, univariate analysis of treatment response.

Figure 14.7. Male versus female, univariate analysis of overall survival.

Figure 14.7

Male versus female, univariate analysis of overall survival.

Figure 14.8. Male versus female, multivariate analysis of overall survival.

Figure 14.8

Male versus female, multivariate analysis of overall survival.

Figure 14.9. Liver involvement, univariate analysis of treatment response.

Figure 14.9

Liver involvement, univariate analysis of treatment response.

Figure 14.10. Liver involvement, univariate analysis of overall survival.

Figure 14.10

Liver involvement, univariate analysis of overall survival.

Figure 14.11. Liver involvement, multivariate analysis of overall survival.

Figure 14.11

Liver involvement, multivariate analysis of overall survival.

Figure 14.12. Lymph node involvement, univariate analysis of treatment response.

Figure 14.12

Lymph node involvement, univariate analysis of treatment response.

Figure 14.13. Lymph node involvement, univariate analysis of overall survival.

Figure 14.13

Lymph node involvement, univariate analysis of overall survival.

Figure 14.14. Lymph node involvement, multivariate analysis of overall survival.

Figure 14.14

Lymph node involvement, multivariate analysis of overall survival.

Figure 14.15. Lung metastases, univariate analysis of treatment response.

Figure 14.15

Lung metastases, univariate analysis of treatment response.

Figure 14.16. Lung metastases, univariate analysis of overall survival.

Figure 14.16

Lung metastases, univariate analysis of overall survival.

Figure 14.17. Elevated serum LDH, univariate analysis of treatment response.

Figure 14.17

Elevated serum LDH, univariate analysis of treatment response.

Figure 14.18. Elevated serum LDH, univariate analysis of overall survival.

Figure 14.18

Elevated serum LDH, univariate analysis of overall survival.

Figure 14.19. Elevated serum LDH, multivariate analysis of overall survival.

Figure 14.19

Elevated serum LDH, multivariate analysis of overall survival.

Figure 14.20. Performance status, univariate analysis of treatment response.

Figure 14.20

Performance status, univariate analysis of treatment response.

Figure 14.21. Performance status, univariate analysis of overall survival.

Figure 14.21

Performance status, univariate analysis of overall survival.

Figure 14.22. Performance status, multivariate analysis of overall survival.

Figure 14.22

Performance status, multivariate analysis of overall survival.

Figure 14.23. Age, univariate analysis of treatment response.

Figure 14.23

Age, univariate analysis of treatment response.

Figure 14.24. Age, univariate analysis of overall survival.

Figure 14.24

Age, univariate analysis of overall survival.

Figure 14.25. Age, multivariate analysis of overall survival.

Figure 14.25

Age, multivariate analysis of overall survival.

Figure 14.26. Number of metastatic sites, univariate analysis of treatment response.

Figure 14.26

Number of metastatic sites, univariate analysis of treatment response.

Figure 14.27. Number of metastatic sites, univariate analysis of overall survival.

Figure 14.27

Number of metastatic sites, univariate analysis of overall survival.

Figure 14.28. Number of metastatic sites, multivariate analysis of overall survival.

Figure 14.28

Number of metastatic sites, multivariate analysis of overall survival.

Figure 14.29. Peritoneal involvement, univariate analysis of treatment response.

Figure 14.29

Peritoneal involvement, univariate analysis of treatment response.

Figure 14.30. Peritoneal involvement, univariate analysis of overall survival.

Figure 14.30

Peritoneal involvement, univariate analysis of overall survival.

Figure 14.31. Peritoneal involvement, multivariate analysis of overall survival.

Figure 14.31

Peritoneal involvement, multivariate analysis of overall survival.

Lactate dehydrogenase (LDH)

Elevated serum LDH was an adverse prognostic factor for overall survival on univariate analysis. Elevated LDH was an independent prognostic factor in five of the nine studies that considered it in multivariate analysis. In these five studies patients with elevated serum LDH had almost twice the risk of death of those with normal serum LDH levels, HR=1.94 [95% C.I. 1.54 to 2.44].

Elevated serum LDH did not significantly affect response to platinum based chemotherapy, RR = 0.98 [0.68 to 1.41], however 95% confidence intervals were wide and included both appreciable benefit and harm.

Serum albumin

Low serum albumin was an independent adverse prognostic factor for overall survival in all three studies that considered it (Assersoh et al 2003; Seve et al 2006a and Munoz et al 2004). Munoz et al 2004 reported that patients with low serum albumin were at greatly increased risk of death, HR = 4.31 [95% C.I. 1.56 to 11.85]. Seve et al (2006a) also found low serum albumin to be an independent risk factor, HR = 2.70 [95% C.I. 1.79 to 4.07]

Serum alkaline phosphatase

Elevated serum alkaline phosphatase was an independent adverse prognostic factor for overall survival in three of the nine studies that examined it in multivariate analysis.

Performance status

Studies of performance status divided people into groups of good performance status and poor performance status. Some studies defined good performance status as 0 on the WHO/ECOG scale, while others defined it as 0 to 1 on the WHO/ECOG scale. Poor PS was everything else. Good performance status (however defined) was a favourable prognostic factor for overall survival in nine of the ten studies that analysed it in multivariate analysis, The pooled hazard ratio in these nine studies was 0.62 [95% C.I. 0.53 to 0.73].

Patients with good performance status were more likely to respond to chemotherapy, RR = 1.60 [1.09 to 2.35] on univariate analysis.

Number of metastatic sites

Studies divided patients into two groups according to the number of metastatic sites. Typically patients with either one or one to two metastatic sites were compared with everyone else. Fewer metastatic sites was a favourable prognostic factor for overall survival, HR = 0.82 [95% C.I. 0.73 to 0.92] on multivariate analysis. Patients with fewer sites were more likely to respond to chemotherapy, RR = 1.64 [95% C.I. 1.18 to 2.29] on univariate analysis.

Age

Studies split patients into two age groups, the cut-point defining older and younger varied between studies from 50 years to 65 years. In chemotherapy series younger age was not a prognostic factor for treatment response or overall survival. In univariate analysis from series of patients not selected by treatment, however, younger age was a favourable prognostic factor for overall survival HR = 0.69 [0.58 to 0.81]. Multivariate analyses suggested age was not an independent prognostic factor.

Histology

Studies were typically restricted to patients with adenocarcinoma, poorly differentiated carcinoma or undifferentiated carcinoma. On univariate analysis adenocarcinoma histology was an adverse prognostic factor for treatment response, RR=0.71 [0.59 to 0.86], and overall survival, HR = 1.32 [1.18 to 1.47]. Multivariate analyses, however, suggested adenocarcinoma histology was not an independent prognostic factor.

Poorly differentiated adenocarcinoma or poorly differentiated carcinoma histology was an positive prognostic factor for treatment response, RR = 1.44 [1.16 to 1.78], and overall survival, HR = 0.78 [0.67 to 0.91].

Evidence from two studies (Van der Gaast el al 1990; Pavlidis et al, 1992), suggests that patients with undifferentiated carcinoma are more than twice as likely to respond to platinum based chemotherapy than patients with other histology. The relative risk for response to treatment was 2.10 [95% C.I. 1.21 to 3.66]

Liver metastases

People with liver metastases tended to have poorer overall survival than people without. On multivariate analysis seven of the twelve studies that considered it found liver metastases to be an adverse prognostic factor for survival. The pooled hazard ratio in these seven studies was 1.40 [95% C.I. 1.24 to 1.57].

The presence of liver metastases was the factor most strongly associated with lack of response to chemotherapy, RR = 0.56 [0.45 to 0.69]

Lung metastases

The presence of lung metastases was an adverse prognostic factor for overall survival, HR=1.40 [1.24 to 1.57] on univariate analysis. It was unlikely that presence of lung metastases was an independent prognostic factor, however, as no studies retained this factor in their multivariate models.

People with lung metastases were also less likely to respond to chemotherapy, RR = 0.70 [0.53 to 0.93].

Peritoneal metastases

Presentation with peritoneal metastases was a favourable prognostic factor for treatment response, RR = 1.45 [95% C.I. 1,12 to 1.88]. There was imprecision and inconsistency in the estimate of the effect on peritoneal metastases on overall survival, and it was unclear whether peritoneal metastasis was a prognostic factor for overall survival.

Lymph node metastases

Lymph node metastases were a independent favourable prognostic factor for overall survival in only two of the nine studies that considered it. The presence of lymph node metastases was the factor most strongly associated with response to chemotherapy, RR = 2.68 [1.94 to 3.70].

Prognostic models

Prognostic models aim to classify patients into risk groups for overall survival and could be used as decision aids in treatment decisions (see Table 4). These models are developed using clinical data from group of patients (the development cohort) but need to be tested in an independent set of patients to confirm their validity.

Table 14.4. Multivariate prognostic models for overall survival.

Table 14.4

Multivariate prognostic models for overall survival.

Culine et al (2002)

Culine et al (2002) developed a prognostic model to classify patients with CUP into high and low risk groups for death from any cause, using two prognostic factors: performance status and serum LDH. In the group of patients used to develop the model the median survival in high and low risk groups was 4 months and 12 months respectively. In an independent set of patients used to validate the model the median survival in high and low risk groups was 7 and 12 months respectively. The model of Culine et al was validated by Van de Wouw et al (2004) who reported median survival of 1 month and 6.5 months median survival for the high and low risk groups in their cohort. Similarly Yonemori et al (2006) reported median survivals of 10 and 21 months for the high and low risk groups using the Culine model (P=0.003). Munoz et al (2008), however, failed to demonstrate a significant difference in the overall survival of the three risk groups in their cohort of patients with CUP.

Van der Gaast et al (1995)

Van der Gaast et al (1995) developed a model for patients with undifferentiated cancer of unknown primary using two prognostic factors: performance status and serum alkaline phosphatase. The median survival of high and intermediate risk groups was 4 and 10 months respectively. Median survival was not reached in the low risk group. Yonemori et al (2006) validated the model of Van der Gast, reporting median survival in the high, intermediate and low risk groups of 20, 12 and 7 months respectively (P not reported).

Ponce Lorenzo (2007)

Ponce Lorenzo (2007) developed a prognostic model to classify patients into three risk groups on the basis of performance status and presence of liver metastases. Munoz et al (2008) challenged this model, after testing it in their CUP cohort, claiming that it failed to discriminate between low and intermediate risk groups well enough. Unsuprisingly the model of Munoz et al (2008), using serum albumin and performance status, performed better in their own cohort (probably because it was developed using the same patients).

Seve et al (2006a)

Seve et al (2006a) reported a prognostic model to divide patients with CUP into high and low risk groups for death from any cause using serum albumin and the presence of liver metastases. The model was validated by the authors in an independent set of patients, with median survival of 3 months and 13 months in the high and low risk groups respectively (P<0.0001). Seve et al (2006a) suggested that the model of Culine et al (2002) was less powerful than their own, in this validation set: using Culine’s model median survival in the high and low risk groups was 4 and 13 months respectively (P=0.07).

Trivanovic et al (2009)

Trivanovic et al (2009) reported a prognostic model to classify patients into three risk groups using the following adverse prognostic factors: elevated LDH, QTc prolongation, liver mets, PS 2 or more, anaemia, age 63 years or more. The model has not been validated.

Hess et al (1999)

Hess et al (1999) used classification and regression tree (CART) analysis to put patients into one of ten risk groups. Their CART model incorporates: presence of liver, bone, adrenal, lymph node and pleural metastases, neuroendocrine histology, age, number of metastatic sites and adenocarcinoma histology. The authors note that validation studies are particularly important for CART models as their structure is highly dependent on the development cohort. No validation studies were found for this model.

Change in management and avoidance of inappropriate treatment

None of the studies reported change in management or avoidance of inappropriate treatment on the basis of prognostic factors

References
  • Abbruzzese JL, Abbruzzese MC, Hess KR, Raber MN, Lenzi R, Frost P. Unknown primary carcinoma: natural history and prognostic factors in 657 consecutive patients. Journal of Clinical Oncology. 1994;12(6):1272–80. [PubMed: 8201389]
  • Al-Kubaisy W. Metastatic Carcinoma of Unknown Origin Treatment with Vinorelbine; Gemcetabine and Methotrexate. Journal of the Bahrain Medical Society. 2003;15(4):199–203.
  • Alberts AS, Falkson G, Falkson HC, van der Merwe MP. Treatment and prognosis of metastatic carcinoma of unknown primary: analysis of 100 patients. Medical & Pediatric Oncology. 1989;17(3):188–92. [PubMed: 2747591]
  • Assersohn L, Norman AR, Cunningham D, Iveson T, Seymour M, Hickish T, et al. A randomised study of protracted venous infusion of 5-fluorouracil (5-FU) with or without bolus mitomycin C (MMC) in patients with carcinoma of unknown primary.[see comment). European Journal of Cancer. 2003;39(8):1121–8. [PubMed: 12736112]
  • Beldi D, Jereczek-Fossa BA, D’Onofrio A, Gambaro G, Fiore MR, Pia F, et al. Role of radiotherapy in the treatment of cervical lymph node metastases from an unknown primary site: retrospective analysis of 113 patients. International Journal of Radiation Oncology, Biology, Physics. 2007;69(4):1051–8. [PubMed: 17716824]
  • Berry W, Elkordy M, O’Rourke M, Khan M, Asmar L. Results of a phase II study of weekly paclitaxel plus carboplatin in advanced carcinoma of unknown primary origin: a reasonable regimen for the community-based clinic? Cancer Investigation. 2007;25(1):27–31. [PubMed: 17364554]
  • Briasoulis E, Tsavaris N, Fountzilas G, Athanasiadis A, Kosmidis P, Bafaloukos D, et al. Combination regimen with carboplatin, epirubicin and etoposide in metastatic carcinomas of unknown primary site: A Hellenic Co-Operative Oncology Group Phase II Study. Oncology. 1998;55(5):426–30. [PubMed: 9732220]
  • Briasoulis E, Kalofonos H, Bafaloukos D, Samantas E, Fountzilas G, Xiros N, et al. Carboplatin plus paclitaxel in unknown primary carcinoma: a phase II Hellenic Cooperative Oncology Group Study. Journal of Clinical Oncology. 2000;18(17):3101–7. [PubMed: 10963638]
  • Briasoulis E, Fountzilas G, Bamias A, Dimopoulos MA, Xiros N, Aravantinos G, et al. Multicenter phase-II trial of irinotecan plus oxaliplatin [IROX regimen] in patients with poor-prognosis cancer of unknown primary: a hellenic cooperative oncology group study. Cancer Chemotherapy & Pharmacology. 2008;62(2):277–84. [PubMed: 17901952]
  • Culine S, Fabbro M, Ychou M, Romieu G, Cupissol D, Pujol H. Chemotherapy in carcinomas of unknown primary site: A high-dose intensity policy. Annals of Oncology. 1999;10(5):569–75. [PubMed: 10416007]
  • Culine S, Kramar A, Saghatchian M, Bugat R, Lesimple T, Lortholary A, et al. Development and validation of a prognostic model to predict the length of survival in patients with carcinomas of an unknown primary site. Journal of Clinical Oncology. 2002;20(24):4679–83. [PubMed: 12488413]
  • Falkson CI, Cohen GL. Mitomycin C, epirubicin and cisplatin versus mitomycin C alone as therapy for carcinoma of unknown primary origin. Oncology. 1998;55(2):116–21. [PubMed: 9499185]
  • Farrugia DC, Norman AR, Nicolson MC, Gore M, Bolodeoku EO, Webb A, et al. Unknown primary carcinoma: Randomised studies are needed to identify optimal treatments and their benefits. European Journal of Cancer. 1996;32A(13):2256–61. [PubMed: 9038607]
  • Greco FA, Hainsworth JD, Yardley DA, Burris HA III, Erland JB, Rodriguez GI, et al. Sequential paclitaxel/carboplatin/etoposide (PCE) followed by irinotecan/gemcitabine for patients (pts) with carcinoma of unknown primary site (CUP) a Minnie Pearl Cancer Research Network phase II trial. Proceedings of the American Society of Clinical Oncology. 2002;21 abstr 642.
  • Hainsworth JD, Johnson DH, Greco FA. Cisplatin-Based Combination Chemotherapy in the Treatment of Poorly Differentiated Carcinoma and Poorly Differentiated Adenocarcinoma of Unknown Primary Site - Results of A 12-Year Experience. Journal of Clinical Oncology. 1992;10(6):912–22. [PubMed: 1375284]
  • Hainsworth JD, Erland JB, Kalman LA, Schreeder MT, Greco FA. Carcinoma of unknown primary site: treatment with 1-hour paclitaxel, carboplatin, and extended-schedule etoposide. Journal of Clinical Oncology. 1997;15(6):2385–93. [see comment] [PubMed: 9196154]
  • Hainsworth JD, Fizazi K. Treatment for patients with unknown primary cancer and favorable prognostic factors. Seminars in Oncology. 2009;36(1):44–51. [PubMed: 19179187]
  • Hauswald H. Predictive factors in patients with cervical lymph node metastases in unknown primary tumours. Strahlentherapie und Onkologie. 2007;183:90.
  • Hauswald H, Lindel K, Rochet N, Debus J, Harms W. Surgery with complete resection improves survival in radiooncologically treated patients with cervical lymph node metastases from cancer of unknown primary. Strahlentherapie und Onkologie. 2008;184(3):150–6. [PubMed: 18330511]
  • Hess KR, Abbruzzese MC, Lenzi R, Raber MN, Abbruzzese JL. Classification and Regression Tree Analysis of 1000 Consecutive Patients with Unknown Primary Carcinoma. Clinical Cancer Research. 1999;5(11):3403–10. [PubMed: 10589751]
  • Jentsch-Ullrich K, Leuner S, Kahl C, Arland R, Florschutz A, Franke A, et al. Prognostic factors for treatment results in patients with carcinoma unknown primary site (CUPS). Cancer Journal. 1998;11(4):196–200.
  • Kambhu SA, Kelsen DP, Fiore J, Niedzwiecki D, Chapman D, Vinciguerra V, et al. Metastatic Adenocarcinomas of Unknown Primary Site - Prognostic Variables and Treatment Results. American Journal of Clinical Oncology-Cancer Clinical Trials. 1990;13(1):55–60. [PubMed: 2106257]
  • Karapetis CS. Epirubicin, cisplatin, and prolonged or brief infusional 5-fluorouracil in the treatment of carcinoma of unknown primary site. Medical Oncology. 2001;18(1):23–32. [PubMed: 11778966]
  • Lorenzo JP, Huerta AS, Beveridge RD, Ortiz AG, Aparisi FA, Kanonnikoff TF, et al. Carcinoma of unknown primary site: development in a single institution of a prognostic model based on clinical and serum variables. Clinical & Translational Oncology. 2007;9(7):452–8. [PubMed: 17652059]
  • Luke C, Koczwara B, Karapetis C, Pittman K, Price T, Kotasek D, et al. Exploring the epidemiological characteristics of cancers of unknown primary site in an Australian population: implications for research and clinical care. Australian & New Zealand Journal of Public Health. 2008;32(4):383–9. [PubMed: 18782405]
  • Macdonald AG, Nicolson MC, Samuel LM, Hutcheon AW, Ahmed FY. A phase II study of mitomycin C, cisplatin and continuous infusion 5-fluorouracil (MCF) in the treatment of patients with carcinoma of unknown primary site. British Journal of Cancer. 2002;86(8):1238–42. [PMC free article: PMC2375343] [PubMed: 11953879]
  • Munoz A. [[Prognostic and predictive factors of patients with cancer of unknown origin treated with a paclitaxel-based chemotherapy]]. Medicina Clinica. 2004;122(6):216–8. [Spanish] [PubMed: 15012889]
  • Munoz A, Fuente N, Rubio I, Ferreiro J, Martinez-Bueno A, Lopez-Vivanco G. Prognostic factors in cancer of unknown primary site. Clinical & Translational Oncology. 2008;10(1):64–5. [comment] [PubMed: 18208796]
  • Pasterz R, Savaraj N, Burgess M. Prognostic factors in metastatic carcinoma of unknown primary. Journal of Clinical Oncology. 1986;4(11):1652–7. [PubMed: 3095502]
  • Pavlidis N, Kosmidis P, Skarlos D, Briassoulis E, Beer M, Theoharis D, et al. Subsets of tumors responsive to cisplatin or carboplatin combinations in patients with carcinoma of unknown primary site. A Hellenic Cooperative Oncology Group Study. Annals of Oncology. 1992;3(8):631–4. [PubMed: 1450045]
  • Pentheroudakis G, Briasoulis E, Karavassilis V, Fountzilas G, Xeros N, Samelis G, et al. Chemotherapy for patients with two favourable subsets of unknown primary carcinoma: Active, but how effective? Acta Oncologica. 2005;44(2):155–60. [PubMed: 15788295]
  • Pentheroudakis G, Briasoulis E, Kalofonos H, Fountzilas G, Economopoulos T, Samelis G, et al. Docetaxel and carboplatin combination chemotherapy as outpatient palliative therapy in carcinoma of unknown primary: A multicentre Hellenic Cooperative Oncology Group phase II study. Acta Oncologica. 2008;47(6):1148–55. [PubMed: 18607872]
  • Piga A, Gesuita R, Catalano V, Nortilli R, Cetto G, Cardillo F, et al. Identification of clinical prognostic factors in patients with unknown primary tumors treated with a platinum-based combination. Oncology. 2005;69(2):135–44. [PubMed: 16127284]
  • Pittman KB. Gemcitabine and carboplatin in carcinoma of unknown primary site: A phase 2 Adelaide Cancer Trials and Education Collaborative study. British Journal of Cancer. 2006;95(10):1309–13. [PMC free article: PMC2360587] [PubMed: 17088914]
  • Ponce Lorenzo J, Segura Huerta A, Diaz Beveridge R, Gimenez Ortiz A, Aparisi Aparisi F, Fleitas Kanonnikoff T, et al. Carcinoma of unknown primary site: development in a single institution of a prognostic model based on clinical and serum variables.[see comment). Clinical & Translational Oncology. 2007;9(7):452–8. [PubMed: 17652059]
  • Saghatchian M, Fizazi K, Borel C, Ducreux M, Ruffie P, Le Chevalier T, et al. Carcinoma of an unknown primary site: a chemotherapy strategy based on histological differentiation--results of a prospective study.[see comment). Annals of Oncology. 2001;12(4):535–40. [PubMed: 11398889]
  • Schneider BJ, El-Rayes B, Muler JH, Philip PA, Kalemkerian GP, Griffith KA, et al. Phase II trial of carboplatin, gemcitabine, and capecitabine in patients with carcinoma of unknown primary site. Cancer. 2007;110(4):770–5. [PubMed: 17594717]
  • Seve P, Sawyer M, Hanson J, Broussolle C, Dumontet C, Mackey JR. The influence of comorbidities, age, and performance status on the prognosis and treatment of patients with metastatic carcinomas of unknown primary site: a population-based study. Cancer. 2006;106(9):2058–66. [PubMed: 16583433]
  • Seve P, Ray-Coquard I, Trillet-Lenoir V, Sawyer M, Hanson J, Broussolle C, et al. Low serum albumin levels and liver metastasis are powerful prognostic markers for survival in patients with carcinomas of unknown primary site. Cancer. 2006;107(11):2698–705. [PubMed: 17063500]
  • Seve P, Mackey J, Sawyer M, Lesimple T, de La Fouchardiere C, Broussolle C, et al. Impact of clinical practice guidelines on the management for carcinomas of unknown primary site: a controlled “before-after” study. Bulletin du Cancer. 2009;96(4):E7–17. [PubMed: 19435692]
  • Sulkes A, Uziely B, Isacson R, Brufman G, Biran S. Combination chemotherapy in metastatic tumors of unknown origin. 5-Fluorouracil, adriamycin and mitomycin C for adenocarcinomas and adriamycin, vinblastine and mitomycin C for anaplastic carcinomas. Israel Journal of Medical Sciences. 1988;24(9–10):604–10. [PubMed: 3204009]
  • Sumi H, Itoh K, Onozawa Y, Shigeoka Y, Kodama K, Ishizawa K, et al. Treatable subsets in cancer of unknown primary origin. Japanese Journal of Cancer Research. 2001;92(6):704–9. [PMC free article: PMC5926750] [PubMed: 11429061]
  • Trivanovic D, Petkovic M, Stimac D. New prognostic index to predict survival in patients with cancer of unknown primary site with unfavourable prognosis. Clinical Oncology (Royal College of Radiologists). 2009;21(1):43–8. [PubMed: 18976894]
  • van de Wouw AJ, Jansen RL, Griffioen AW, Hillen HF. Clinical and immunohistochemical analysis of patients with unknown primary tumour. A search for prognostic factors in UPT. Anticancer Research. 2004;24(1):297–301. [PubMed: 15015611]
  • Van Der Gaast A, Verweij J, Planting AS, Hop WC, Stoter G. Simple prognostic model to predict survival in patients with undifferentiated carcinoma of unknown primary site. Journal of Clinical Oncology. 1995;13(7):1720–5. [PubMed: 7541451]
  • Voog E, Merrouche Y, Trillet-Lenoir V, Lasset C, Peaud PY, Rebattu P, et al. Multicentric phase II study of cisplatin and etoposide in patients with metastatic carcinoma of unknown primary. American Journal of Clinical Oncology. 2000;23(6):614–6. [PubMed: 11202809]
  • Wagener DJT, Demulder PHM, Burghouts JT, Croles JJ. Phase-Ii Trial of Cisplatin for Adenocarcinoma of Unknown Primary Site. European Journal of Cancer. 1991;27(6):755–7. [PubMed: 1829919]
  • Warner E, Goel R, Chang J, Chow W, Verma S, Dancey J, et al. A multicentre phase II study of carboplatin and prolonged oral etoposide in the treatment of cancer of unknown primary site (CUPS). British Journal of Cancer. 1998;77(12):2376–80. [PMC free article: PMC2150411] [PubMed: 9649162]
  • Woods RL. A randomized study of two combination-chemotherapy regimens. New England Journal of Medicine. 1980;303(2):87–9. [PubMed: 6991941]
  • Yonemori K, Ando M, Shibata T, Katsumata N, Matsumoto K, Yamanaka Y, et al. Tumor-marker analysis and verification of prognostic models in patients with cancer of unknown primary, receiving platinum-based combination chemotherapy. Journal of Cancer Research & Clinical Oncology. 2006;132(10):635–42. [PubMed: 16791594]
  • Tierney JF, Stewart LA, Ghersi D, Burdett S, Sydes MR. Practical methods for incorporating summary time-to-event data into meta analysis. Trials. 2007;8(16) [PMC free article: PMC1920534] [PubMed: 17555582]

Characteristics of included studies

Abbruzzese-1994

Al-Kubaisy-2003

Alberts-1989

Assersohn-2003

Berry-2007

Briasoulis-1998

Briasoulis-2000

Briasoulis-2008

Culine-1999

Culine-2002

Falkson-1998

Farrugia-1996

Greco-2004

Hainsworth-1992

Hainsworth-1997

Hess-1999

Jentsch-Ullrich-1998a

Kambhu-1990

Karapetis-2001

Macdonald-2002

Munoz-2004

Munoz-2008

Pasterz-1986

Pentheroudakis-2008

Piga-2005

Pittman-2006

Ponce-2007

Saghatchian-2001

Schneider-2007

Seve-2006

Seve-2006a

Sulkes-1988

Sumi-2001

Trivanovic-2009

van-de-Wouw-2004

Van-Der-Gaast-1995

Voog-2000

Wagener-1991

Warner-1998

Woods-1980

Yonemori-2006

References for included studies

Abbruzzese 1994.
Abbruzzese JL, Abbruzzese MC, Hess KR, Raber MN, Lenzi R, Frost P. Unknown primary carcinoma: natural history and prognostic factors in 657 consecutive patients. Journal of Clinical Oncology. 1994;12(6):1272–80. [PubMed: 8201389]
Al Kubaisy 2003.
Al-Kubaisy W. Metastatic Carcinoma of Unknown Origin Treatment with Vinorelbine; Gemcetabine and Methotrexate. Journal of the Bahrain Medical Society. 2003;15(4):199–203.
Alberts 1989.
Alberts AS, Falkson G, Falkson HC, van der Merwe MP. Treatment and prognosis of metastatic carcinoma of unknown primary: analysis of 100 patients. Medical & Pediatric Oncology. 1989;17(3):188–92. [PubMed: 2747591]
Assersohn 2003.
Assersohn L, Norman AR, Cunningham D, Iveson T, Seymour M, Hickish T, et al. A randomised study of protracted venous infusion of 5-fluorouracil (5-FU) with or without bolus mitomycin C (MMC) in patients with carcinoma of unknown primary. European Journal of Cancer. 2003;39(8):1121–8. [see comment] [PubMed: 12736112]
Beldi 2007.
Beldi D, Jereczek-Fossa BA, D’Onofrio A, Gambaro G, Fiore MR, Pia F, et al. Role of radiotherapy in the treatment of cervical lymph node metastases from an unknown primary site: retrospective analysis of 113 patients. International Journal of Radiation Oncology, Biology, Physics. 2007;69(4):1051–8. [PubMed: 17716824]
Berry 2007.
Berry W, Elkordy M, O’Rourke M, Khan M, Asmar L. Results of a phase II study of weekly paclitaxel plus carboplatin in advanced carcinoma of unknown primary origin: a reasonable regimen for the community-based clinic. Cancer Investigation. 2007;25(1):27–31. [PubMed: 17364554]
Briasoulis 1998.
Briasoulis E, Tsavaris N, Fountzilas G, Athanasiadis A, Kosmidis P, Bafaloukos D, et al. Combination regimen with carboplatin, epirubicin and etoposide in metastatic carcinomas of unknown primary site: A Hellenic Co-Operative Oncology Group Phase II Study. Oncology. 1998;55(5):426–30. [PubMed: 9732220]
Briasoulis 2000.
Briasoulis E, Kalofonos H, Bafaloukos D, Samantas E, Fountzilas G, Xiros N, et al. Carboplatin plus paclitaxel in unknown primary carcinoma: a phase II Hellenic Cooperative Oncology Group Study. Journal of Clinical Oncology. 2000;18(17):3101–7. [PubMed: 10963638]
Briasoulis 2008.
Briasoulis E, Fountzilas G, Bamias A, Dimopoulos MA, Xiros N, Aravantinos G, et al. Multicenter phase-II trial of irinotecan plus oxaliplatin [IROX regimen] in patients with poor-prognosis cancer of unknown primary: a hellenic cooperative oncology group study. Cancer Chemotherapy & Pharmacology. 2008;62(2):277–84. [PubMed: 17901952]
Culine 1999.
Culine S, Fabbro M, Ychou M, Romieu G, Cupissol D, Pujol H. Chemotherapy in carcinomas of unknown primary site: A high-dose intensity policy. Annals of Oncology. 1999;10(5):569–75. [PubMed: 10416007]
Culine 2002.
Culine S, Kramar A, Saghatchian M, Bugat R, Lesimple T, Lortholary A, et al. Development and validation of a prognostic model to predict the length of survival in patients with carcinomas of an unknown primary site. Journal of Clinical Oncology. 2002;20(24):4679–83. [PubMed: 12488413]
Falkson 1998.
Falkson CI, Cohen GL. Mitomycin C, epirubicin and cisplatin versus mitomycin C alone as therapy for carcinoma of unknown primary origin. Oncology. 1998;55(2):116–21. [PubMed: 9499185]
Farrugia 1996.
Farrugia DC, Norman AR, Nicolson MC, Gore M, Bolodeoku EO, Webb A, et al. Unknown primary carcinoma: Randomised studies are needed to identify optimal treatments and their benefits. European Journal of Cancer. 1996;32A(13):2256–61. [PubMed: 9038607]
Greco 2004.
Greco FA, Hainsworth JD, Yardley DA, Burris HA III, Erland JB, Rodriguez GI, et al. Sequential paclitaxel/carboplatin/etoposide (PCE) followed by irinotecan/gemcitabine for patients (pts) with carcinoma of unknown primary site (CUP) a Minnie Pearl Cancer Research Network phase II trial. Proceedings of the American Society of Clinical Oncology. 2002;21 abstr 642.
Hainsworth 1992.
Hainsworth JD, Johnson DH, Greco FA. Cisplatin-Based Combination Chemotherapy in the Treatment of Poorly Differentiated Carcinoma and Poorly Differentiated Adenocarcinoma of Unknown Primary Site - Results of A 12-Year Experience. Journal of Clinical Oncology. 1992;10(6):912–22. [PubMed: 1375284]
Hainsworth 1997.
Hainsworth JD, Erland JB, Kalman LA, Schreeder MT, Greco FA. Carcinoma of unknown primary site: treatment with 1-hour paclitaxel, carboplatin, and extended-schedule etoposide. Journal of Clinical Oncology. 1997;15(6):2385–93. [see comment] [PubMed: 9196154]
Hainsworth 2009.
Hainsworth JD, Fizazi K. Treatment for patients with unknown primary cancer and favorable prognostic factors. Seminars in Oncology. 2009;36(1):44–51. [PubMed: 19179187]
Hauswald 2007.
Hauswald H. Predictive factors in patients with cervical lymph node metastases in unknown primary tumours. Strahlentherapie und Onkologie. 2007;183:90.
Hauswald 2008.
Hauswald H, Lindel K, Rochet N, Debus J, Harms W. Surgery with complete resection improves survival in radiooncologically treated patients with cervical lymph node metastases from cancer of unknown primary. Strahlentherapie und Onkologie. 2008;184(3):150–6. [PubMed: 18330511]
Hess 1999.
Hess KR, Abbruzzese MC, Lenzi R, Raber MN, Abbruzzese JL. Classification and Regression Tree Analysis of 1000 Consecutive Patients with Unknown Primary Carcinoma. Clinical Cancer Research. 1999;5(11):3403–10. [PubMed: 10589751]
Jentsch-Ullrich 1998A.
Jentsch-Ullrich K, Leuner S, Kahl C, Arland R, Florschutz A, Franke A, et al. Prognostic factors for treatment results in patients with carcinoma unknown primary site (CUPS). Cancer Journal. 1998;11(4):196–200.
Kambhu 1990.
Kambhu SA, Kelsen DP, Fiore J, Niedzwiecki D, Chapman D, Vinciguerra V, et al. Metastatic Adenocarcinomas of Unknown Primary Site - Prognostic Variables and Treatment Results. American Journal of Clinical Oncology-Cancer Clinical Trials. 1990;13(1):55–60. [PubMed: 2106257]
Karapetis 2001.
Karapetis CS. Epirubicin, cisplatin, and prolonged or brief infusional 5-fluorouracil in the treatment of carcinoma of unknown primary site. Medical Oncology. 2001;18(1):23–32. [PubMed: 11778966]
Lorenzo 2007.
Lorenzo JP, Huerta AS, Beveridge RD, Ortiz AG, Aparisi FA, Kanonnikoff TF, et al. Carcinoma of unknown primary site: development in a single institution of a prognostic model based on clinical and serum variables. Clinical & Translational Oncology. 2007;9(7):452–8. [PubMed: 17652059]
Luke 2008.
Luke C, Koczwara B, Karapetis C, Pittman K, Price T, Kotasek D, et al. Exploring the epidemiological characteristics of cancers of unknown primary site in an Australian population: implications for research and clinical care. Australian & New Zealand Journal of Public Health. 2008;32(4):383–9. [PubMed: 18782405]
Macdonald 2002.
Macdonald AG, Nicolson MC, Samuel LM, Hutcheon AW, Ahmed FY. A phase II study of mitomycin C, cisplatin and continuous infusion 5-fluorouracil (MCF) in the treatment of patients with carcinoma of unknown primary site. British Journal of Cancer. 2002;86(8):1238–42. [PMC free article: PMC2375343] [PubMed: 11953879]
Munoz 2004.
Munoz A. [[Prognostic and predictive factors of patients with cancer of unknown origin treated with a paclitaxel-based chemotherapy]]. Medicina Clinica. 2004;122(6):216–8. [Spanish] [PubMed: 15012889]
Munoz 2008.
Munoz A, Fuente N, Rubio I, Ferreiro J, Martinez-Bueno A, Lopez-Vivanco G. Prognostic factors in cancer of unknown primary site. Clinical & Translational Oncology. 2008;10(1):64–5. [comment] [PubMed: 18208796]
Pasterz 1986.
Pasterz R, Savaraj N, Burgess M. Prognostic factors in metastatic carcinoma of unknown primary. Journal of Clinical Oncology. 1986;4(11):1652–7. [PubMed: 3095502]
Pavlidis 1992.
Pavlidis N, Kosmidis P, Skarlos D, Briassoulis E, Beer M, Theoharis D, et al. Subsets of tumors responsive to cisplatin or carboplatin combinations in patients with carcinoma of unknown primary site. A Hellenic Cooperative Oncology Group. Study Annals of Oncology. 1992;3(8):631–4. [PubMed: 1450045]
Pentheroudakis 2005.
Pentheroudakis G, Briasoulis E, Karavassilis V, Fountzilas G, Xeros N, Samelis G, et al. Chemotherapy for patients with two favourable subsets of unknown primary carcinoma: Active, but how effective. Acta Oncologica. 2005;44(2):155–60. [PubMed: 15788295]
Pentheroudakis 2008.
Pentheroudakis G, Briasoulis E, Kalofonos H, Fountzilas G, Economopoulos T, Samelis G, et al. Docetaxel and carboplatin combination chemotherapy as outpatient palliative therapy in carcinoma of unknown primary: A multicentre Hellenic Cooperative Oncology Group phase II study. Acta Oncologica. 2008;47(6):1148–55. [PubMed: 18607872]
Piga 2005.
Piga A, Gesuita R, Catalano V, Nortilli R, Cetto G, Cardillo F, et al. Identification of clinical prognostic factors in patients with unknown primary tumors treated with a platinum-based combination. Oncology. 2005;69(2):135–44. [PubMed: 16127284]
Pittman 2006.
Pittman KB. Gemcitabine and carboplatin in carcinoma of unknown primary site: A phase 2 Adelaide Cancer Trials and Education Collaborative study. British Journal of Cancer. 2006;95(10):1309–13. [PMC free article: PMC2360587] [PubMed: 17088914]
Ponce 2007.
Ponce Lorenzo J, Segura Huerta A, Diaz Beveridge R, Gimenez Ortiz A, Aparisi Aparisi F, Fleitas Kanonnikoff T, et al. Carcinoma of unknown primary site: development in a single institution of a prognostic model based on clinical and serum variables. Clinical & Translational Oncology. 2007;9(7):452–8. [see comment] [PubMed: 17652059]
Saghatchian 2001.
Saghatchian M, Fizazi K, Borel C, Ducreux M, Ruffie P, Le Chevalier T, et al. Carcinoma of an unknown primary site: a chemotherapy strategy based on histological differentiation--results of a prospective study. Annals of Oncology. 2001;12(4):535–40. [see comment] [PubMed: 11398889]
Schneider 2007.
Schneider BJ, El-Rayes B, Muler JH, Philip PA, Kalemkerian GP, Griffith KA, et al. Phase II trial of carboplatin, gemcitabine, and capecitabine in patients with carcinoma of unknown primary site. Cancer. 2007;110(4):770–5. [PubMed: 17594717]
Seve 2006.
Seve P, Sawyer M, Hanson J, Broussolle C, Dumontet C, Mackey JR. The influence of comorbidities, age, and performance status on the prognosis and treatment of patients with metastatic carcinomas of unknown primary site: a population-based study. Cancer. 2006;106(9):2058–66. [PubMed: 16583433]
Seve 2006a.
Seve P, Ray-Coquard I, Trillet-Lenoir V, Sawyer M, Hanson J, Broussolle C, et al. Low serum albumin levels and liver metastasis are powerful prognostic markers for survival in patients with carcinomas of unknown primary site. Cancer. 2006;107(11):2698–705. [PubMed: 17063500]
Seve 2009.
Seve P, Mackey J, Sawyer M, Lesimple T, de La Fouchardiere C, Broussolle C, et al. Impact of clinical practice guidelines on the management for carcinomas of unknown primary site: a controlled “before-after” study. Bulletin du Cancer. 2009;96(4):E7–17. [PubMed: 19435692]
Sulkes 1988.
Sulkes A, Uziely B, Isacson R, Brufman G, Biran S. Combination chemotherapy in metastatic tumors of unknown origin. 5-Fluorouracil, adriamycin and mitomycin C for adenocarcinomas and adriamycin, vinblastine and mitomycin C for anaplastic carcinomas. Israel Journal of Medical Sciences. 1988;24(9–10):604–10. [PubMed: 3204009]
Sumi 2001.
Sumi H, Itoh K, Onozawa Y, Shigeoka Y, Kodama K, Ishizawa K, et al. Treatable subsets in cancer of unknown primary origin. Japanese Journal of Cancer Research. 2001;92(6):704–9. [PMC free article: PMC5926750] [PubMed: 11429061]
Trivanovic 2009.
Trivanovic D, Petkovic M, Stimac D. New prognostic index to predict survival in patients with cancer of unknown primary site with unfavourable prognosis. Clinical Oncology (Royal College of Radiologists). 2009;21(1):43–8. [PubMed: 18976894]
van de Wouw 2004.
van de Wouw AJ, Jansen RL, Griffioen AW, Hillen HF. Clinical and immunohistochemical analysis of patients with unknown primary tumour. A search for prognostic factors in UPT. Anticancer Research. 2004;24(1):297–301. [PubMed: 15015611]
van der Gaast 1995.
Van Der Gaast A, Verweij J, Planting AS, Hop WC, Stoter G. Simple prognostic model to predict survival in patients with undifferentiated carcinoma of unknown primary site. Journal of Clinical Oncology. 1995;13(7):1720–5. [PubMed: 7541451]
Voog 2000.
Voog E, Merrouche Y, Trillet-Lenoir V, Lasset C, Peaud PY, Rebattu P, et al. Multicentric phase II study of cisplatin and etoposide in patients with metastatic carcinoma of unknown primary. American Journal of Clinical Oncology. 2000;23(6):614–6. [PubMed: 11202809]
Wagener 1991.
Wagener DJT, Demulder PHM, Burghouts JT, Croles JJ. Phase-Ii Trial of Cisplatin for Adenocarcinoma of Unknown Primary Site. European Journal of Cancer. 1991;27(6):755–7. [PubMed: 1829919]
Warner 1998.
Warner E, Goel R, Chang J, Chow W, Verma S, Dancey J, et al. A multicentre phase II study of carboplatin and prolonged oral etoposide in the treatment of cancer of unknown primary site (CUPS). British Journal of Cancer. 1998;77(12):2376–80. [PMC free article: PMC2150411] [PubMed: 9649162]
Woods 1980.
Woods RL. A randomized study of two combination-chemotherapy regimens. New England Journal of Medicine. 1980;303(2):87–9. [PubMed: 6991941]
Yonemori 2006.
Yonemori K, Ando M, Shibata T, Katsumata N, Matsumoto K, Yamanaka Y, et al. Tumor-marker analysis and verification of prognostic models in patients with cancer of unknown primary, receiving platinum-based combination chemotherapy. Journal of Cancer Research & Clinical Oncology. 2006;132(10):635–42. [PubMed: 16791594]

15. Decision aids for people with cancer of unknown primary

Last updated: 30/10/2009.

Short summary

Decision aids are designed to help people understand options, consider the personal importance of harms and benefits and to take part in the decision making process (O’Connor et al 2009).

There was an absence of published decision aids for people with cancer of unknown primary.

There is good evidence, from randomised trials, that decision aids are useful when patients need to make diagnostic or treatment decisions in cancer. When compared with usual care, decision aids improved people’s knowledge of their options and reduced difficulty with decision making.

Rationale

Rationale for asking this question needs to be written, including the key decisions faced by patients with CUP.

Methods

Study types

There was no restriction on study design.

Participants

People with confirmed cancer of unknown primary.

Interventions

Decision aids, such as pamphlets and videos, that describe treatment or diagnostic options. The comparison is usual care, with no decision aids. According to O’Connor et al (2009) decision aids are designed to help people make specific and deliberative choices among options (including the status quo) by providing (at the minimum) information on the options and outcomes relevant to a person’s health status.

Outcomes

Patient satisfaction with decision making, decisional conflict, knowledge acquisition and anxiety.

Search results

Description of included studies

The literature search found no studies of decision aids for people with confirmed cancer of unknown primary. Several studies developed prognostic models for patients with CUP (Seve et al 2006; Trivanovic et al 2009; Culine et al 2002; Penel et al, 2009). These could form the basis of a decision aid for treatment decisions, but they have not yet been evaluated as such.

A high level search of MEDLINE for systematic reviews of decision aids in patients with cancer identified one recent systematic review (O’Brien et al 2009) and a Cochrane review of decision aids for people facing health treatment or screening decisions (O’Connor et al 2009).

These reviews included no studies in patients with cancer of unknown primary, but many of the trials addressed similar decisions to those faced by patients with cancer of unknown primary. There were 22 randomised trials of screening for cancer, where people decided whether to proceed with a diagnostic test after considering the potential harms and benefits of diagnosis. Similarly there were also trials of decision aids for treatment options, when there was no obvious best treatment choice and patients had to consider the personal importance of the various harms and benefits when choosing.

Study quality

Both the included systematic reviews were of high quality.

Evidence summary

Knowledge Acquisition

Both reviews (O’Brien et al 2009 and O’Connor et al 2009) found that decision aids significantly improved people’s knowledge of their options when compared with usual care. O’Connor et al estimated the magnitude of this improvement as approximately 15% (95% CI 12% to 19%; where knowledge was rated on a scale of 0 to 100%).

Decisional conflict

Decisional conflict is a composite measure that includes the patient’s comfort with decisional making in terms of how well informed they feel, the clarity of their values, how supported they feel in the decision making process, and their level of uncertainty (O’Brien et al 2009).

Both reviews (O’Brien et al 2009 and O’Connor et al 2009) found that decision aids reduced people’s decisional conflict when compared with usual care. O’Connor et al estimated the magnitude of this reduction as approximately 8% (95% CI 5% to 12%; where decisional conflict was rated on a scale of 0 to 100%).

Satisfaction and Anxiety

Neither review found an effect of decision aids on patients satisfaction with their decision or on their levels of anxiety. It is plausible that information about treatment outcomes and harms could increase anxiety in some cases.

References
  • O’Brien MA, Whelan TJ, Villas is Keever M, Gafni A, Charles C, Roberts R, Schiff S, Cai W. Are Cancer-Related Decision Aids Effective? A Systematic Review and Meta-Analysis. Journal of Clinical Oncology. 2009;27(6):974–985. [PubMed: 19124808]
  • O’Connor AM, Bennett CL, Stacey D, Barry M, Col NF, Eden KB, Entwistle VA, Fiset V, Holmes-Rovner M, Khangura S, Llewellyn-Thomas H, Rovner D. Decision aids for people facing health treatment or screening decisions. Cochrane Database of Systematic Reviews. 2009;(3) Art. No.: CD001431. [PubMed: 19588325] [CrossRef]
  • Culine S, Kramar A, Saghatchian M, Bugat R, Lesimple T, Lortholary A, et al. Development and validation of a prognostic model to predict the length of survival in patients with carcinomas of an unknown primary site. Journal of Clinical Oncology. 2002;20(24):4679–83. [PubMed: 12488413]
  • Penel N, Negrier S, Ray-Coquard I, Ferte C, Devos P, Hollebecque A, Sawyer MB, Adenis A, Seve P. Development and validation of a bedside score to predict early death in cancer of unknown primary patients. PLoS ONE. 2009;4(8):e6483. [PMC free article: PMC2715134] [PubMed: 19649260]
  • Seve P, Sawyer M, Hanson J, Broussolle C, Dumontet C, Mackey JR. The influence of comorbidities, age, and performance status on the prognosis and treatment of patients with metastatic carcinomas of unknown primary site - A population-based study. Cancer. 2006;106(9):2058–66. [PubMed: 16583433]
  • Trivanovic D, Petkovic M, Stimac D. New prognostic index to predict survival in patients with cancer of unknown primary site with unfavourable prognosis. Clinical Oncology. 2009;21(1):43–8. [PubMed: 18976894]

Characteristics of included studies

O-Brien-2009

O-Connor-2009

References for included studies

O Brien 2009.
O’Brien MA, Whelan TJ, Villasis Keever M, Gafni A, Charles C, Roberts R, Schiff S, Cai W. Are Cancer-Related Decision Aids Effective? A Systematic Review and Meta-Analysis. Journal of Clinical Oncology. 2009;27(6):974–985. [PubMed: 19124808]
O Connor 2009.
O’Connor AM, Bennett CL, Stacey D, Barry M, Col NF, Eden KB, Entwistle VA, Fiset V, Holmes-Rovner M, Khangura S, Llewellyn-Thomas H, Rovner D. Decision aids for people facing health treatment or screening decisions. Cochrane Database of Systematic Reviews. 2009;(3) Art. No.: CD001431. [PubMed: 19588325] [CrossRef]
Copyright © 2010, National Collaborating Centre for Cancer.

No part of this publication may be reproduced, stored or transmitted in any form or by any means, without the prior written permission of the publisher or, in the case of reprographic reproduction, in accordance with the terms of licenses issued by the Copyright Licensing Agency in the UK. Enquiries concerning reproduction outside the terms stated here should be sent to the publisher at the UK address printed on this page.

The use of registered names, trademarks etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore for general use.

Bookshelf ID: NBK82139
Contents