The value of routine follow-up after treatment for head and neck cancer. A National Survey from DAHANCA

Abstract

Background. The post-treatment follow-up is well-integrated in the oncologic care tradition, based on the risk of developing recurrent disease or new primary tumors in treated patients. Furthermore, follow-up serves as an opportunity to monitor treatment effects and to provide clinical care of side effects. In this study we measured the activity and effectiveness of routine follow-up in head and neck cancer and assessed the value of follow-up from the perspectives of both physicians and the patients. Patients and methods. During a period of six weeks a prospective national cross section cohort of 619 patients attending regular follow-up were enrolled. All patients had received intended curative treatment for head and neck cancer and all were followed according to DAHANCA guidelines. Data were collected by the physician filling in a registration form containing chosen key parameters and patients filling in a validated questionnaire. Results. The majority (91%) of the 619 visits was planned, and 75% of all visits included either tumor or treatment-related problems. Suspicion of recurrent disease led to further diagnostic work-up in 80 visits (13%). A total of 29 recurrences were found, and of these seven (25%) were asymptomatic, i.e. the “number needed to see” to detect one asymptomatic recurrence was 99. Treatment-related normal-tissue problems were addressed in 72% of all visits, and among these 18% required intervention. Although the majority of problems (either suspicion of recurrent disease or late effects) occurred within a few years after treatment, 39% of patients seen after three years also had problems. The majority of patients (97%) expressed satisfaction with the planned follow-up. Conclusion. Only few relapses are found in asymptomatic patients at routine follow-up, with one silent recurrence detected per 99 visits. However, head and neck cancer survivors have a substantial need for management of sequelae. In this context, a centralized routine follow-up may still be worthwhile.

Post-treatment follow-up (FU) is well integrated in the oncologic care tradition, and in head and neck cancer (HNC), regular FU is generally regarded as indispensable [1], since patients are at risk of developing both recurrent tumors and new primary cancer. Secondly, FU serves as an opportunity to monitor effects of treatment and to provide patients with appropriate clinical care for side effects.

There are no existing generally accepted standards for FU of patients treated for cancer [2]. In Denmark all patients with HNC are followed according to DAHANCA guidelines and both treatment and FU are centralized and government funded. This involves a five-year post-treatment FU strategy with patients seen more frequently the first two years (every third month in the first year and every fourth month in the second) and then every sixth month in the third and fourth year and finally after the completed fifth year seen once. The structure of this FU program is designed to identify early recurrences, which is known to be more frequent in the first two years, and 80% of all recurrences occur within three years [3].

In 1996 a study conducted at the head and neck oncology centers at Aarhus University Hospital and Rigshospitalet in Copenhagen showed that a specialist FU was justified for evaluation of treatment results and to treat and advise patients with moderate to severe radiation-related morbidity [4]. Increasing economic and other resource pressures have led to a review of whether current practice and polices are cost-effective. Arguments against regular FU exist for other cancer types, where randomized studies have not been able to demonstrate any differences in survival or quality of life. Investigations have shown that only few relapses are found at the FU visit of asymptomatic patients contrary to symptomatic patients [5–7], indicating that symptom directed FU could be more effective.

The objectives of this study were to measure and document the activity and effectiveness of a routine FU clinic, and to assess the value of FU from the perspectives of both physicians and patients.

Patients and methods

The study was conducted during a six-week period in all five Danish HNC centers between January and February 2012. A total of 619 consecutive patients attending the regular out-patient clinic were prospectively included. All patients had been treated with curative intent for their primary tumor. Patients with previously known persistent or recurrent active disease were not included. Patients attending two or more times during the period were only enrolled at the first visit. At FU all patients underwent a full clinical examination of the head and neck, oral cavity and pharynx by either a specialist oncologist or an otolaryngologist.

The adherence to the recommended centralized FU program and hence the completeness of the current cohort was tested in a consecutive series of all 324 HN patients treated with curative radiotherapy in Aarhus 2006–2008. The patient cohort was identified from the departmental radiotherapy information system, which routinely capture all patients treated. Of these 324 patients, only five patients (1.5%) did not adhere to the routine FU program, indicating that the routine FU clinics capture the vast majority of all HNC patients.

Non-appearance at scheduled appointments was examined in a sample of the 215 patients included in Aarhus. It was found that although 11% cancelled or did not show up on the scheduled FU visit, all were seen on a later time point within the next weeks.

The data were collected using a registration form concerning age, tumor site, treatment year, treatment modality and whether the visit was regular or extraordinary (due to either patient or physician demand), patient symptoms and status at previous visit in the department and physician's findings. Consequences of the FU such as referral to x-ray, MR/CT/US scan, endoscopies, other investigations, counseling, prescription of medicine and biopsy were also recorded. Finally, registration of treatment related late effects and the handling of it, such as counseling, referral to physiotherapist/occupational therapist/speech therapist/dentist/otolaryngologist etc.

All patients were also asked to fill in a questionnaire after the FU visit. The questionnaire has previous been validated in 2303 patients at 13 centers in eight European countries [8]. An independent native translator having Danish as his mother-tongue translated the original questionnaire from English to Danish. Back-translation was used as a quality control step to ensure that the original meaning of the concepts was derived. The back-translation was conducted by an independent native translator with English as her mother tongue and without knowledge of the questionnaire beforehand. A pretest investigated the level of comprehensibility and cognitive equivalence of the translation among 20 patients seen in the FU clinic. The patients filled in the questionnaire right after the visit. The first author briefly interviewed the patient regarding problems with the questions. This was to ensure a simple and acceptable language.

Data were recorded in the DAHANCA database and statistical analysis was performed using Stata 11 software package. The actual values of the endpoints were estimated using 2´2 contingency tables and χ²-test, using a significance level of 5%. The primary endpoints were the number of asymptomatic recurrences relative to all recurrences and visits, the proportion of morbidity related visits relative to all visits, and the proportion of visits resulting in referrals for morbidity management relative to visits with morbidity.

The project was discussed with the Regional Ethics Committee, which concluded that the study was a quality assurance project, and as such it did not need formal approval according to Danish regulations.

Results

The basic characteristics of the 619 enrolled patients are shown in Table I. Most visits were from patients coming for routine FU (91%). Despite this 75% of all visits included either tumor or treatment related problems. In total, recurrent disease or a new primary cancer was suspected in 13% and led to further diagnostic work-up. Figure 1 illustrates the frequency of investigations and actions taken in the 99 patients where cancer was suspected.

Figure 1. Suspicion of recurrence in 99 patients. Investigations and actions (n = 133).

Table I. Patient and tumor characteristics.

Variables	n	%
Gender
Male	422	68
Female	197	32
Age
Median	66
Range	(16–95)
Distribution of tumor site
Nasopharynx	12	2
Oropharynx	199	32
Hypopharynx	36	6
Larynx (glottic)	106	17
Larynx (others)	38	6
Oral cavity	137	22
Sino-nasal	36	6
Salivary gland	27	4
Other	28	5
Treatment
Radiotherapy	253	41
Chemo-radiotherapy	166	27
Postoperative radiotherapy	83	13
Surgery	107	17
Other	10	2
Time since primary treatment
< 1 year	152	25
1–2 years	179	29
2–3 years	116	19
3–4 years	85	14
4–5 years	45	7
> 5 years	42	7
Status at former visit
No recurrence	518	86
Suspicion of recurrent disease	26	4
Treated for recurrence, now disease free	61	10
Patient seen by
Otolaryngologist	55	9
Oncologist	256	41
Joint clinic	304	50
Type of visit:
Planned visit	565	91
Extra visit	44	7
Other	10	2

Recurrent disease or a new cancer was diagnosed in 29 patients in 619 visits (5%). The distribution of recurrences was 52% T-site recurrence, 31% N-site, 10% M-site, and 7% had a new primary cancer diagnosed. The highest incidence of events was found in patients with cancer of the oral cavity (28%), followed by glottic larynx (21%), oro-pharyngeal (17%), other cancer sites (17%), sino-nasal (14%) and the sites with lowest incidence of events were carcinoma of the infra- and supra glottic larynx (3%) and naso-pharynx (0%). The distribution and cumulative percentage of recurrences as a function of time is shown in Table II. Of all the events, 33% were diagnosed in the first year after curative treatment of the tumor. The cumulative percentage after two years of FU was 55% and after three years 69% and after four years 83%. The two new primaries were diagnosed the third and fourth year after initial primary treatment, respectively.

Table II. Distribution of verified recurrences of disease or new primaries over the follow-up period.

Years since primary treatment	Number of visits in group	Recurrence	New primary	Cum. %
0–1	152	9	0	31%
1–2	179	7	0	55%
2–3	116	3	1	69%
3–4	85	3	1	83%
4–5	45	1	0	86%
5+	42	4	0	100%
Total	619	27	2

The registration form contained identical fields for the patient's and the clinician's perception of possible problems regarding recurrence of cancer, respectively (Table III). Recurrent disease was suspected by the physician in 98 cases. Of the 69 patients who feared a recurrence, 17 patients could be reassured after the actual examination and 31 after further diagnostic work-up. In the 52 situations where both the patient and the physician suspected recurrence, a cancer was confirmed in 22 (42%), compared to 15% when only the physician suspected recurrence. The probability of the physician having “no suspicion” of recurrence in an asymptomatic patient was 91%. On the other hand, the probability of “suspicion of recurrence” occurring in a patient who had reported symptoms was 75%; and 42% of them were subsequently verified. The risk of having a verified recurrence being an asymptomatic patient attending FU was 1.3%. This implies that the “numbers needed to see” to detect recurrent disease in an asymptomatic patient is 99.

Table III. Cross-tabulation of the perception of potential recurrence by the physicians and patients, respectively.

		Physician suspects recurrence
		No	Yes	Verified/ All suspected
Patient suspects recurrence	No	482 (81%)	47 (8%)	7/47 (15%)
	Yes	17 (3%)	52 (9%)	22/52 (42%)

Of the 29 verified recurrences, 22 patients also had symptoms. In 17 of the 29 recurrences, patients were found to be candidates for salvage treatment with curative intent. All seven patients with asymptomatic recurrences underwent salvage treatment, compared to nine of 22 patients with symptomatic recurrence (p < 0.05). Nineteen of all symptomatic recurrences were presented at a regular routine FU.

Data regarding late morbidity and the management of it is listed in Table IV. Treatment related normal tissue problems were addressed in 445 visits corresponding to 72%, and among these 18% required an intervention. This intervention consisted in 216 of 445 cases (49%) of physician counseling only. Counseling was primarily regarding artificial saliva, smoking cessation, exercise and nutrition. Referral to management of late morbidity elsewhere occurred in 24% (109 of 445) of visits.

Table IV. Distribution of morbidity and the applied interventions.

Variables	Number of incidences	%
Morbidity
Caries	45	7
Dysphagia/Eating problems	121	20
Low state of nutrition	45	7
Hearing trouble/Tinnitus	10	2
Hoarseness	110	18
Osteonecrosis	12	2
Denture	82	13
Taste disorder	38	6
Pain	57	9
Trismus	26	4
Respiratory problems	5	1
Xerostomia/Viscous mucus	205	33
Other	114	18
Management of Morbidity
None	189	31
Counseling	216	35
Physiotherapist	7	1
Occupational Therapist	4	1
Speech Therapist	11	2
Dietician	1	< 1
Social Worker	2	< 1
Rehabilitation	4	1
Referral to further Investigations	6	1
Medical Prescription	21	3
Dentist	24	4
Jaw Surgeon	8	1
ENT Physician	20	3
Other	24	4

Figure 2 shows the data of tumor and treatment-related problems in relation to years since primary treatment. The problems were more marked in the first years after primary treatment. Patients seen after the third year reported fewer symptoms related to treatment. Late morbidity was recorded in 122 of the 295 visits after the third year (41%). Referral for management of morbidity problems was in 60% of cases done in early FU (0–3 years) and in 40% in late FU. Counseling was equally distributed between early and late FU, as was the case for patients receiving neither counseling nor management of their problems.

Figure 2. Tumor and treatment-related problems, related to years after primary treatment.

The questionnaire was completed by 532 of 619 (70%) of all invited patients. For the question “Did you report any health problems to your doctor today?” 65% (347 of 532) answered positively and 30% of them answered to the question “What do you think is causing your problem?” that the problem was related to the treatment, 29% meant that the problem was related to their illness, 24% that it was related to both illness and treatment and the rest (18%) ticked off at the box that other things were causing their problem.

When asked the question “Did you or your relative(s) ask any questions today?” about half answered positively and half negatively. For the patients answering positively 34% stated that the question was about “Illness”, “Treatment” in 17%, “Both Treatment and Illness” in 30% and 20% ticked off that the question was about other things. To the question, “Were the question answered during the consultation?”, 96% responded that they felt that their question was answered. In answer to the question “Was there anything else you would have liked to discuss with the doctor?”, 95% replied no. Finally 97% answered yes to the question “Are you satisfied with the FU plan the doctor presented today?”

Discussion

The current study was undertaken to evaluate the outcome of the current FU practice for HNC in Denmark, and form the basis for discussions on potential future recommendations.

Patients treated for HNC routinely undergo surveillance to achieve a number of goals. The main expectation of frequent clinical FU is to allow clinicians to identify asymptomatic and early recurrences that would lead to earlier and more successful salvage treatment and thereby improve outcomes. Based on literature review and their own experience an international panel recently recommended that patients are followed in the clinic every two to three months for the first two years, every three to six months for years three and four, every six months during the fifth year and annually thereafter [2]. However, the panel also acknowledged that only low level evidence exists to support this recommendation.

A study of 905 patients from the Danish DAHANCA material showed a difference in five-year survival between patients with primary loco-regional control and patients with loco-regional failure to be 98% and 29%, respectively [9]. In the subgroup of patients with loco-regional failure who later obtained loco-regional control after salvage therapy had a five-year survival as high as 77%. The value of and the cost-effectiveness with early diagnosis of recurrent disease has been illustrated in a large prospective study of 661 patients [10]. They found recurrence of disease in 220 cases dispersed over 7813 visits. The rate of diagnosis of recurrent disease was thus one in 36 visits and the following cure rate was one in 113 visits. The best results were as expected seen with primary disease in larynx or oral cavity, where surgical removal of tumor was possible. Only 39% of the recurrences were asymptomatic and found by physical examination alone. The remaining 61% all had alarming symptoms of recurrent disease. These figures correlated well with the former study by Grau et al. from 1996. Here, the authors found 11 asymptomatic recurrences with physical examination alone corresponding to 32% [4].

In the present material “silent” recurrent disease was found in seven patients corresponding to 24% of all cases of diagnosed recurrences. This is somewhat less than the result of Boysen et al. [10] and Grau et al. [4], but corresponds well with the experience of Agrawal et al. [6]. They found in an analysis of 3645 prospectively collected visits recurrence of disease or new primary cancer in 5%, and 65% of the diagnosed cancer cases were judged to be candidates for salvage therapy. Of the 180 recurrences or new primaries there were 142 patients (79%) who had presented new symptoms at the FU visits. In fact, 40% of all patients presenting symptoms were diagnosed with cancer. It is noteworthy that the physician´s examination generally took place at the time of the regular scheduled FU visit, despite the appearance of new symptoms between visits. Asymptomatic patients diagnosed with cancer constituted 21% and the authors stated that the risk of having a cancer diagnosed in the absence of symptoms was 1.2% [6].

In an analysis of 1039 prospectively collected patients Kothari et al. showed that suspicion of recurrence occurred in 10% of patients seen routinely [5]. This figure increased to 68% for the subset of patients who had requested an extra appointment. Only 3% of asymptomatic patients attending routine FU were suspected of having a recurrence. Finally, they reported that 54% of all recurrences happened within the first year. The figures of the present study are somewhat lower, as 31% had recurrent disease after one year [5].

Overgaard et al. 1986 showed that 80% of recurrences occur within the first three years of treatment. This is generally the reason to recommend a more frequent surveillance in the first years after completed primary treatment and thereby accommodate this predisposition of early recurrence of disease [2]. A Dutch study analyzing the FU of 428 patients concluded that routine patient FU regimes were indispensable, but the length of FU should be tumor dependent. They found that recurrent disease was verified in one of 34 asymptomatic patients and in one of 2.7 in symptomatic patients. However, the reason for their recommendation was that recurrent disease in asymptomatic patients resulted in better post-recurrence survival than in symptomatic recurrences (58 months versus 32 months) [1]. Making the FU regimen tumor dependent is also supported by Lester et al. Their recommendation is based on an analysis of 676 patients diagnosed with squamous cell carcinoma of the head and neck. They found 105 cases with recurrence of disease and 20 cases with second primary tumors. By calculating the mean of “the time to new cancer event” and stratifying for primary tumor location, they state that primary cancers of the larynx should be followed for seven years and oropharyngeal and hypopharyngeal primary tumors for three years [11].

Another substantial reason for centralized surveillance of HNC patients is management of the specific, persistent side effects after surgery and/or irradiation. Late effects after radiotherapy such as fibrosis, edema, xerostomia, dental problems and nutritional deficits are common [2,12–14]. Late effects can develop several years after completed primary treatment. In a Norwegian study by Björdal et al. of 204 head and neck cancer survivors, a large proportion had physical problems, considerably reduced quality of life and emotional problems [15]. Grau et al. showed that 61% of FU visits included one or more problems related to either tumor or treatment-related morbidity. About 50% of all visits included treatment related morbidity, and 30% had problems that required an intervention. Even three to four years after primary treatment 47% of the patients still had one or more problems [4]. In the current study the numbers were even higher, as 72% of all patients reported symptoms of late effects. Again the incidence were higher in the first years after treatment, but even after the third year 41% still suffered from these conditions. Management besides counseling was necessary in 24% of all cases.

Patients cured for cancer of the head and neck, have a considerably increased risk of developing new primary tumors not only of the head and neck, but also in other regions, in particular, in the aerodigestive tract and bladder [10,16,17]. In a Danish analysis of 478 larynx cancer patients who were followed during a 10-year period after primary treatment, 81 new primary tumors were diagnosed (17%). Several patients with previous T1 larynx cancer later died from a lung cancer. In two Norwegian studies, 9% and 17% new primaries were found within a maximum 10 years of FU after completed primary treatment [10,16]. New primary tumors were rarely diagnosed at scheduled FU, since this visit does not routinely involve pan- endoscopies or chest x-rays. In addition, the overall cost benefit of the additional effort to find a second primary cancer of the lung or esophagus is doubtful, since the treatment results generally are disappointing. Regarding new cancers of the head and neck, the detection rate is better, since 20% were found at patient's routine FU. An early diagnosis will, in these cases, increase the chance of referral of the patient to treatment with curative intent.

Whether routine FU improves early tumor detection and overall survival of patients developing new primary tumor is controversial. Rennemo et al. analyzed a material of 351 new primaries occurring in a cohort of 2063 HNC patients. The median survival after diagnosis was 12 months. The group of patients who developed new primaries had a longer median overall survival compared with all other patients included. They were younger patients with limited initial tumors [16]. This could emphasize the need for secondary prophylactic initiatives, as informing of the risk of continuing smoking and drinking (and recommend cessation) and teaching patients regarding alarm signals, as already done in Danish FU clinics. Nevertheless, our data suggests that considerably more time should be spent on informing patients about the possible symptoms and signs of recurrent cancer disease, as the vast majority of patients with diagnosed symptomatic disease (85%) were attending regular scheduled FU. Patients should be instructed to contact the out-patient department in these situations.

The patient's attitude towards the FU consultation is also of importance. Analysis of this questionnaire revealed that patients in general were satisfied with the FU. Few patients (3%) had complaints or concerns, and these were about long transport distance, increased anxiety, too infrequent visits, and finally changing physicians from time to time. In general FU seems to satisfy the needs of the patients. Boysen et al. state that patients prior to consultations are anxious that a recurrence will be found. If the examination turns out well the patients are relieved and this may induce a feeling of security. This feeling may sometimes be false and lead to neglect of alarming symptoms. Thus, the effect of the FU regimen may sometimes cause a delay in the diagnosis of a recurrence [10]. Kothari et al. conclude from a questionnaire survey that there is a need for a more focused, individualized approach to FU in HNC. They hope that this would result in reduced overbooking and better informed less anxious patients with individually tailored FU plans and easier access to clinical examination [5].

The ideal regime for FU of patients with head and neck cancer continues to be a matter of debate, with very low level of evidence [2]. The data specific for head and neck cancer surveillance are entirely retrospective. Randomized data will be necessary to formulate cost-effective, clinically appropriate FU guidelines. In the existing literature only one study has shown a survival difference in favor of regular FU [1]. There is little evidence for regular FU for improving disease control and/or survival, but surveillance seems to be important for other reasons, including evaluation of efficacy of the initial treatment, detection and potential eradication of any treatment-induced complications, management of additional cancer-related morbidities, provision of psychological and psychosocial support to the patient during rehabilitation and enhancement of the patient-physician relationship [2,18,19]. Development and testing of novel strategies of FU, including nurse-led or ad hoc consults with self-presentation will be of interest. However, it has been argued by some investigators that such alternatives should be used only in combination with planned surveillance and should not replace it [2].

In conclusion, the current study suggests that only few relapses are found in asymptomatic patients at routine FU, with only one silent recurrence detected per 99 FU visits. However, HNC survivors have a substantial need for management of sequelae and rehabilitation, and 75% of all FU visits addressed one or more problems even several years after treatment. In this context, a centralized routine FU for HNC survivors may still be worthwhile.

Acknowledgements

We appreciate the support from CIRRO – The Lundbeck Foundation Center for Interventional Research in Radiation Oncology and The Danish Council for Strategic Research. The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.

Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.