Follow-up care for survivors of lymphoma who have received curative-intent treatment

Methods The medline and embase databases and the Cochrane Database of Systematic Reviews were searched for evidence published between 2000 and August 2015 relating to lymphoma survivorship follow-up. The evidence summary was developed by a Working Group at the request of the Cancer Care Ontario Survivorship and Cancer Imaging programs because of the absence of evidence-based practice documents in Ontario for the follow-up and surveillance of asymptomatic patients with lymphoma in complete remission.


INTRODUCTION
The lymphomas are made up of a large group of neoplasms that arise from the lymphatic system.In 2014, the Leukemia & Lymphoma Society of Canada estimated that there would be 9000 new cases of lymphoma diagnosed in Canada (1000 Hodgkin lymphomas and 8000 non-Hodgkin lymphomas), making lymphoma the sixth most common malignancy in Canada (1).There are many types and subtypes of non-Hodgkin lymphoma.Worldwide, diffuse large B-cell lymphoma (DLBCL) represents the most common subtype of non-Hodgkin lymphoma, accounting for 30 to 40 per cent of all newly diagnosed cases (2).DLBCL and Hodgkin lymphoma (HL) are considered curable with therapies that include chemotherapy, immunotherapy and radiation.
A significant proportion of patients will relapse, typically within the first two years after primary treatment, and many can be treated successfully for cure with salvage chemotherapy and stem cell transplantation.For this reason, surveillance is considered important in this group to detect relapse as early as possible based on the assumption that earlier detection will lead to better outcomes by detecting subclinical disease with a lower tumour burden.
Surveillance testing, which includes physical examination, blood testing and imaging, is currently used to follow patients with DLBCL and HL who are considered to be in remission after treatment to detect recurrence.There is known to be a wide variation in practice, especially in the frequency of imaging, and recent population studies have suggested significant over testing in asymptomatic patients may occur, and not result in improved outcomes.Currently there are no Canadian guidelines which summarize the evidence regarding the type and timing of surveillance testing for asymptomatic patients with DLBCL and HL who have been treated for cure.The intent of this evidence summary is to assess the available evidence on the follow-up of asymptomatic survivors of lymphoma who have received curative-intent treatment.

RESEARCH QUESTIONS
Three research questions were developed to direct the search for available evidence on the follow-up of asymptomatic survivors of lymphoma who have received curative-intent treatment.
1. What clinical activities have been shown to be effective at detecting clinical recurrence or further hematological neoplasm?
2. What is the appropriate frequency and timing for the clinical activities that have been shown to be effective at detecting clinical recurrence or further hematological neoplasm (malignancy)?3. What surveillance procedures have been shown to be effective at detecting therapy-related secondary malignancies following treatment for lymphoma?
TARGET POPULATION All asymptomatic adolescent (≥15 years) and adult survivors of lymphoma who have completed curative-intent treatment and undergo routine follow-up for lymphoma; these patients may or may not be followed by an oncologist.

INTENDED PURPOSE
The purpose of this evidence summary is to report the available data regarding the posttreatment follow-up and surveillance of patients with lymphoma treated with curative intent.

INTENDED USERS
This evidence summary is targeted for all people involved in clinical follow-up of asymptomatic survivors of lymphoma who have received curative-intent treatment including:  Hematologists  Medical and radiation oncologists  Radiologists  Family physicians  Nurses  Administrators and policy makers

METHODS
This evidence summary was developed by a Working Group consisting of one radiation oncologist, two hematologists, one regional primary care lead, two radiologists, one registered nurse, two patient representatives, and a health research methodologist at the request of the Cancer Care Ontario Survivorship programs, from the Clinical Programs and Quality Initiatives, due to the absence of evidence-based practice documents in Ontario for the follow-up and surveillance of asymptomatic patients with lymphoma treated with curative intent.The Working Group was responsible for reviewing the identified evidence and drafting the summary.Information regarding members of the Working Group can be found in Appendix 1.
Conflict-of-interest declarations for all authors and reviewers are summarized in Appendix 2, and were managed in accordance with the PEBC Conflict of Interest Policy.
This evidence review was conducted in two planned stages; a search for systematic reviews followed by a search for primary literature.These stages are described in subsequent sections.

Search for Existing Systematic Reviews
The Cochrane Database of Systematic Reviews was searched from January 2000 to August 2015 using the word "lymphoma".Likewise, the MEDLINE and EMBASE databases were searched using Ovid to identify existing systematic reviews that addressed one or more of the research questions above.Systematic reviews older than five years were considered not relevant, because the main goal of a search for systematic reviews is to identify recent secondary sources covering the primary relevant literature on the topic of interest, the follow-up care for survivors of lymphoma who have received curative-intent treatment.Medical Subject Heading (MeSH) terms related to lymphoma follow-up were combined with text words as a search filter.The full literature strategy used to retrieve potential relevant studies is presented in Appendix 3.
Systematic reviews were included if they met the following criteria: 1.The existing systematic reviews searched for studies evaluating the follow-up care for adult and/or adolescent survivors of lymphoma who have received curativeintent treatment.2. The literature search strategy for the existing review was reproducible and appropriate.3. The existing systematic review reported the sources searched as well as the dates that were searched.
Identified systematic reviews that met the eligibility criteria would be assessed using A Measurement Tool to Assess Systematic review (AMSTAR) (3) to determine whether or not an existing review could be incorporated as part of the evidentiary base.Any identified reviews that did not meet the criteria above, whose AMSTAR assessments indicated important deficiencies in reporting completeness, or that were otherwise not incorporated as part of the evidence base would be reported in the reference list, but not further described or discussed.

Search for Primary Literature
If no existing systematic review or evidence-based practice guideline was identified, or if identified reviews were incomplete or out-of-date, a systematic review of the primary literature was also planned.

Literature Search Strategy
The MEDLINE (Ovid) (1946 through August 31, 2015) and EMBASE (Ovid) (1996 through Week 35, 2015) databases were searched for evidence in August 2015.The search strategy included the MeSH "exp lymphoma" combined with additional terms and text words for the intervention (follow-up) and the population (survivors).The results were limited to English language articles and articles published from 2000 to 2015.The full literature strategy used to retrieve potential relevant studies is presented in Appendix 3.

Study Selection Criteria and Process Inclusion Criteria:
Articles identified in this literature search were eligible for inclusion if they met the following criteria: 1.Primary studies evaluating the the follow-up care for adult and/or adolescent survivors of lymphoma who have received curative-intent treatment.2. Published full-report articles of randomized or nonrandomized studies with an appropriate control group, decision model studies, and single-arm studies with a sample size of at least 100 patients.3. Studies reporting the outcomes of interest such as recurrence/relapse rate, overall survival rate, and relapse-free survival rate.

Exclusion Criteria:
Studies were excluded if they were: 1. Abstracts, letters, case reports, comments, books, notes, or editorial-type publications.
2. Articles published in a language other than English because resources were not available for translation services.3. Single-arm studies with a sample size <100 patients.
A review of the titles and abstracts that resulted from the search was conducted by one reviewer (NV).For those items that warranted full text review, one reviewer (NV) assessed each item independently and consulted with other members of the Working Group whenever there was uncertainty.

Data Extraction and Assessment of Study Quality and Potential for Bias
Data extraction was conducted by one reviewer (NV).All extracted data and information was assessed by a second reviewer (JS), and audited by an independent auditor to verify the accuracy of the information obtained from the studies included in this report.For primary studies, key characteristics, including author, year of publication, study design, study population, sample size, posttreatment follow-up protocol, and median follow-up time were recorded.Outcomes of interest including relapse rate, time to relapse, method of relapse detection and detection rate by follow-up activity, overall survival rate, and relapse-free survival rate were extracted when available.
Any randomized clinical trial would be assessed for quality by examining the following seven criteria: method of randomization, reporting of blinding, power and sample size calculation, length of follow-up, reporting details of the statistical analysis, reporting on withdrawals to treatment and other losses to follow-up, and reporting on the sources of funding for the research.Comparative, nonrandomized, and single-arm evidence would be assessed according to full reporting of the patient selection criteria, the follow-up each patient received, all relevant outcomes, and the source of funding.All authors reviewed and discussed a draft of this report with the aim of assessing the quality of the evidence as a whole, without the use of a scoring system or cut-offs, according to the policy of the PEBC

Search for Existing Systematic Reviews
The search for systematic reviews identified one citation which was retrieved for full text review (4), but the outcomes of interest were not reported on.No other relevant systematic reviews were identified.

Literature Search Results
As presented in Figure 1, out of 1950 titles and abstracts identified in the search of the MEDLINE and EMBASE databases, 1841 appeared potentially eligible on initial review, and 124 of these were verified to be eligible for full-text review.From these, 11 full-report studies were identified that addressed the the follow-up care for adult and/or adolescent survivors of lymphoma who have received curative-intent treatment and reported the outcome of interest.The remaining 113 studies were excluded because they failed to pass the inclusion criteria.Table 1 summarizes the number and types of studies included per research question.

Study and Patient Characteristics
The primary literature search identified 11 nonrandomized retrospective studies assessing the follow-up of asymptomatic survivors of lymphoma who have received curativeintent treatment, and reporting the outcomes of interest: overall survival rate and relapserelated outcomes (relapse detected by different follow-up schedules, symptomatic versus asymptomatic relapses, relapse-free survival rate, median time to relapse, number of imaging tests per relapse detected).The included studies involved patients with diffuse large B-cell lymphoma, classical Hodgkin lymphoma, lymphoid malignancies, and aggressive Hodgkin lymphoma.See Table 2.

Study Design and Quality
The primary literature returned 11 nonrandomized retrospective studies that met the inclusion and exclusion criteria.A description of the study designs and quality of the studies is presented in Appendix 4. Overall, the body of the evidence is limited mainly by its design based on retrospective analysis of electronic medical records, and the relatively small sample size with low number of relapses.The sample size of the included studies ranged from a low of 109 (7) to a high of 1221 in a population-based study comparing the survival rate of patients with lymphoma undergoing different clinical follow-up policies (8).The majority of the studies encompassed patients with non-Hodgkin lymphoma (5,8,(10)(11)(12)14,15); three studies focused on the follow-up of patients with Hodgkin lymphoma (6,7,13), and one study reported on both types of lymphoma (9).The number of relapses ranged from a low of 15 (10) to a high of 163 (15) in patients with non-Hodgkin lymphoma, and from a low of 11 (13) to a high of 42 ( 9) in patients with Hodgkin lymphoma.To compare the outcomes of patients with cHL who underwent routine surveillance imaging vs clinical surveillance.
Adult patients newly diagnosed with cHL at three participating academic tertiary care medical centres, and who have achieved complete remission confirmed by CT and/or PET at the end of the first-line therapy.

Detection of Relapse
Nine studies reported on the follow-up care of asymptomatic survivors of lymphoma who have received curative-intent treatment (5)(6)(7)(9)(10)(11)(12)(13)15).Two studies involving patients with non-Hodgkin lymphoma in complete remission detected a statistically significant difference in number of relapses initially suspected by clinical manifestations (patient-reported symptoms or physical examination) as compared with those initially suspected by imaging before clinical manifestation (10,15).The study reported by Hong et al. (10) assessed the role of routine imaging versus symptom-directed unplanned early outpatient department visits in patients with DLBCL and reported that early visits due to any symptoms or signs have a strong association with the detection of relapse compared with planned visits with or without clinical symptoms or signs (rate: 33% versus 0.5%; p<0.001).Similarly, the study reported by Truong et al. (15) found that patient-reported symptoms led to the detection of the majority of relapses in aggressive non-Hodgkin lymphoma (86% versus 14%; p<0.0001).
Two additional studies in patients with non-Hodgkin lymphoma detected that the proportion of relapses initially suspected by clinical manifestations ranged from a low of 54% (5) to a high of 78% (11,12), and the proportion of those initially suspected by surveillance imaging ranged from a low of 22% (11,12) to a high of 46% (5).
Three studies involved patients with Hodgkin lymphoma (6,7,13).The study reported by Pingali et al. ( 13) compared the incidence of relapse between patients managed with clinical surveillance alone versus those who underwent routine surveillance imaging, and reported that differences between groups were not statistically significant (7.4% versus 3.4%; p=0.39).The two remaining studies reported that the proportion of relapses in patients initially suspected by clinical manifestations ranged from a low of 13% (6) to a high of 64% (7), and the proportion of those initially suspected by surveillance imaging ranged from a low of 8% (6) to a high of 27% (7).
The study reported by Hong et al. (10) found a median times from relapse to death and overall survival times for 11 patients with relapse initially detected by early unplanned visits (clinical manifestations) of 6.7 and 38.3 months, respectively; but determining whether routine imaging can prolong the survival of relapsed patients was not possible due to the small number (n=4) of patients with relapse initially detected by planned visits with or without routine imaging (3 and 1, respectively).Three of these relapses, detected on planned visits with imaging, had times from relapse to death of 5.7, 7.9, and 9.0 months, and overall survival times of 17.1, 18.9, and 50.2 months; the time from relapse to death and overall survival time of the other patient with relapse detected on a planned visit without routine imaging were 7.6 and 51.9 months, respectively.

Time to Relapse
Four of the studies reported on this outcome (6,(12)(13)(14).Only the study reported by Lin et al. (12) detected a significant benefit for patients with first presentation of relapse found by clinical manifestations compared with patients with asymptomatic relapse found by surveillance imaging (mean: 4.5 versus 6.0 months; p=0.042).The study conducted by Thompson et al. (14), reported the median times from diagnosis to relapse in asymptomatic patients to be 19 and 11 months in a cohort of patients from the United States and France, respectively; the median time from diagnosis to relapse in patients with clinical manifestations of relapse was not reported.The study conducted by Dann et al. (6) reported median time to relapse of 8.6 months for both patients undergoing routine clinical follow-up and patients undergoing routine clinical follow-up with routine imaging.Pingali et al (13) reported median times to relapse in patients with Hodgkin lymphoma as 33 and 18 months for relapses initially suspected by clinical manifestations and those initially suspected by imaging, respectively.

Frequency of Imaging
Three of the studies reported on frequency of imaging (6,12,13).Two of these studies found that in patients with Hodgkin lymphoma, routine surveillance imaging was statistically significantly associated with a higher number of scans when compared with clinical surveillance.Dan et al. (6), reported that routine imaging required 47.5 imaging studies to detect a single relapse, compared with 4.7 imaging studies in the clinical follow-up arm.The number of imaging studies required per patient in the routine imaging follow-up arm was 3.9 compared with 0.6 in the clinical follow-up arm (p<0.001).Similarly, the study conducted by Pingali et al. (13) reported a scan rate in the routine surveillance imaging group to be 4.5 times greater than the rate in the clinical surveillance group (0.89 versus 0.21, respectively; p<0.0001); the number of scans performed per relapse detected was 127 in the routine surveillance imaging arm compared with 14.6 scans in the clinical surveillance group.
* Danish Lymphoma Group Registry † Swedish Lymphoma Registry

Question 2: What is the appropriate frequency and timing for the clinical activities that have been shown to be effective at detecting clinical recurrence or further hematological neoplasm (malignancy) in asymptomatic survivors of lymphoma who have received curative-intent treatment?
The literature search did not return any study specifically designed to evaluate the effectiveness of different frequencies and timings of follow-up schedules on asymptomatic survivors of lymphoma who have received curative-intent treatment.However, the nine studies that were discussed while addressing the research question 1 listed the follow-up schedules used by the institutions from which each population was selected and their relationship with relapse detection.Eight of these studies described the follow-up schedules used by single institutions (5)(6)(7)9,10,12,14,15).The study reported by El-Galaly et al (8) described the follow-up schedule of two neighbouring Scandinavian countries with similar health care systems but completely different traditions for routine imaging: Denmark and Sweden.The majority of the studies reported a clinical follow-up every two to three months for the first two years, then every four to six months for the following three years (years 3-5), with annual visits afterwards.Surveillance imaging was mainly performed in cases where relapse was suspected.A full description of the follow-up schedules reported by each study is presented in Table 4.

Question 3: What surveillance procedures have been shown to be effective at detecting therapy-related secondary malignancies in asymptomatic survivors of lymphoma who have received curative-intent treatment?
The literature search did not return any study specifically designed to evaluate followup schedules to detect therapy-related secondary malignancies in asymptomatic survivors of lymphoma who have received curative-intent treatment.Documentation of therapy-related secondary malignancies may be more available in the radiation safety literature rather than in the lymphoma diagnosis and/or follow-up literature.

DISCUSSION
There is accumulating descriptive literature that suggests that patients with lymphoma treated with curative intent who achieve complete remission (CR) may not benefit from routine surveillance with diagnostic imaging.Currently, routine surveillance protocols, often informed by clinical trials protocols and local practice culture, include history, physical examination, blood tests, and imaging.Surveillance investigations are based on the presumption that early detection of recurrence may improve the outcomes of patients in CR because of a higher likelihood of successful response to salvage therapy due to lower clinical burden.It is also recognized that certain therapies may be associated with a predictable incidence of late organ adverse effects, such as heart disease or second cancers, and some routine testing is directed toward monitoring of the development of these complications.In this review, we sought to examine the evidence for surveillance and toxicity screening in this population.
Currently, no Canadian consensus document exists to set out the optimal follow-up care for asymptomatic survivors of lymphoma who have received curative-intent treatment.This evidence summary was framed by three areas of inquiry: clinical activities to detect relapse, the frequency and timing of clinical activities to detect relapse, and activities to detect therapy-related secondary malignancies in survivors of lymphoma.
Eleven retrospective studies were identified that specifically reported on surveillance activities to detect recurrence.Complete remission was mainly defined by CT scan criteria.In the majority of studies, a planned imaging approach, most often with computed tomography (CT) scans, was compared with imaging that was carried out in response to signs and symptoms.The study populations included aggressive histology NHL and Hodgkin lymphoma of stages I through III.There were no prospective comparisons found.In all studies, no significant differences were found between planned versus unplanned visit on survival, our key outcome of interest.Unfortunately, since all nonrandomized studies carry an unclear risk of bias, the quality of the evidence supporting this summary is low.
There is a lack of consistent evidence to support routine imaging surveillance in survivors of lymphoma who have been treated with curative intent, and who are considered to be in remission at the completion of all planned therapy.It was noted in many of the studies that even on the planned surveillance arms of the studies, the majority of relapses were detected in the interval between planned imaging appointments, most often initiated by signs and symptoms reported by the patients.
We also reviewed the clinical visit schedules reported in the trials.In nine studies, the timing of clinical visits was described.We were unable to find neither any studies that compared routine clinical visits with visits only in response to symptom, nor any comparison of the use of routine blood work versus blood work at the discretion of the treating oncology team, and therefore, no clinical visit schedule was described.The majority of the studies reported a clinical follow-up every two to three months for the first two years, then every four to six months for the following three years (years 3-5), with annual visits afterwards.Surveillance imaging was mainly performed in cases of suspected relapse.It is recognized that most relapses will occur in the first two to three years after completion of therapy and this is reflected in a clinical visit pattern that was fairly consistent between studies.This pattern is similar to the 2015 National Comprehensive Cancer Network (NCCN) guideline (16); the follow-up of patients with Hodgkin lymphoma should be mainly based on interim history and physical examination; CT scans are acceptable once during the first 12 months, and should be clinically prompted afterwards.Similarly, the 2015 NCCN guideline (17) for patients with non-Hodgkin lymphoma recommends mainly clinical follow-up with imaging only as clinically indicated for patients with DLBCL stages I and II, and no more often than every six months for

Figure 1 .
Figure 1.Literature search flow diagram of included studies addressing the follow-up care for adult and/or adolescent survivors of lymphoma who have received curative-intent treatment

Table 4 .
Frequency and timing of clinical activities for detecting clinical recurrence or further hematological neoplasm in asymptomatic survivors of lymphoma who have received curative-intent treatment.Clinical visits: history, physical, CBC  First 2 years: Every 2-3 months  Years 3-5: Every 4-6 months  Years 5+: Annually Scans (CT or FDG-PET/CT) -At discretion of the attending physician Dann et al. (6) [Israel, New Zealand] Arm 1: Clinical surveillance  First 3 years: Every 3-4 months  Year 3-5: Every 6 months  Imaging: Only when clinical findings suspicious for relapse Arm 2: Imaging surveillance Clinical surveillance and imaging as follows:  First 2 years: Every 6 months  Year 3: Once First 2 years: Every 3-4 months  Years 3-5: Every 6 months  Years 5+: Annually Surveillance scans (PET/CT or CT) -routinely performed  First year: Every 4 months  Year 2: Every 6 months  Years 3-5: Annually Cheah et al. (5) [Australia] Surveillance imaging (PET/CT)  First 2 years: Every 6 months  Years 3-5: Annually Lin et al. (12) [Taiwan] Clinical visits and laboratory analysis (blood count with a differential, serum lactate dehydrogenase, and serum beta 2-microglobulin)  First 2 years: Every 1-3 months Surveillance Imaging (CT) -Routinely Performed (head, neck, chest, abdomen, and pelvis)  First 2 years: Every 3-6 months or when clinically indicated  Years 3-5: Annually or when clinically indicated Goldschmidt et al. (9) [Israel] Clinical visits  First 2 years: Every 3-4 months  Years 3-5: Every 6 months  Years 5+: Annually Surveillance imaging (CT, PET, or PET/CT)  First 2 years: Every 6 months  End of year 3: Once Dryver et al. (7) [Canada] Clinical visits -Clinical assessment (history and physical), a chest x-ray, CBC  First 2 years: Every 3 months  Years 3-5: Every 6 months  Years 5+: Annually Surveillance scans -At the discretion of the treating physician  X-ray: Conducted during the clinical visits Symptom assessment, clinical examination, blood test  First 2 years: Every 3 months  Years 3-5: Every 6 month  First 2 years: Every 3-4 months  Year 3: Every 6 months  Years 4-5: Annually Surveillance Scans (CT) -Neck, thorax, abdomen  First 2 years: Every 6 months According to a survey among attending lymphoma specialists from 6 large Danish hematology centres, all hematologists prescribed routine CT scans during the first 2 years of follow-up:  CT every 6 months for 2 years: 94%  CT annually for 1 or 2 years: 6% Prescribed CT after the second year of follow-up: 15% Only if relapse is clinically suspected In Sweden, routine imaging for DLBCL in CR is discouraged by the national guidelines, and in a survey of the 10 major hematology/oncology centres covering >90% of the total Swedish lymphoma population, all centres reported adherence to the guidelines Thompson et al.First 3 years: Every 6 months  Years 3+: Annually  First 2 years: Every 3 months  Years 3-5: Every 6 months  Years 5+: Annually Surveillance Scans (CT) Not reported  First year: At 6 and 12 months (frequency of CT scan adapted to the initial stage and prognostic score) CBC (complete blood count); CR (complete remission); DLBCL (diffuse large B-cell lymphoma); FDG-PET/CT (fluorodeoxyglucose-positron emission tomography/computed tomography); CT (computed tomography); PET (positron emission tomography).