Open questions in watchful waiting for follicular lymphoma

Authors


Nastoupil et al (2016) recently reported the outcome following watchful waiting (WW) in a large retrospective cohort of patients with stage II–IV follicular lymphoma (FL) from the National LymphoCare Study (NLCS), showing a similar overall survival (OS) to that of patients treated immediately after diagnosis. This was shown in several trials before the advent of immunochemotherapy (Brice et al, 1997; Ardeshna et al, 2003) but only in one trial since then (Ardeshna et al, 2014) (Table 1).

Table 1. Freedom from treatment and survival with watchful waiting for follicular lymphoma
Study, yearDesign N Treatment strategyTTT/PFSaTT2T/FFTF/PFSAbOS
  1. WW, watchful waiting; OS, overall survival; PFS, progression-free survival; TTT, time to treatment; PFSA, progression-free survival from first active lymphoma treatment; FFTF, freedom from treatment failure; R-CHT, rituximab-chemotherapy; RCVP, rituximab, cyclophosphamide, vincristine, prednisolone; R-CHOP, rituximab, cyclophosphamide, doxorubicin, vincristine, prednisolone; NR, not reached.

  2. a

    TTT (Time to start of new treatment, where WW is considered first-line treatment) applies to Brice et al (1997), Ardeshna et al (2003, 2014) and Solal-Céligny et al (2012). PFS (time from diagnosis until disease progression or death from any cause) applies to Nastoupil et al (2016). El-Galaly et al (2015) defined PFS as time from diagnosis until lymphoma treatment or death from any cause.

  3. b

    TT2T (time to second line treatment) applies to Ardeshna et al (2003), FFTF (time from diagnosis until disease progression, salvage treatment or death from any cause where first line treatment is not considered an event in the WW group) applies to Solal-Céligny et al (2012) and PFSA (time from initiation of first active lymphoma treatment until disease progression or death from any cause) to Nastoupil et al (2016).

  4. c

    Also includes lymphocytic well-differentiated lymphoma.

Brice et al (1997)Randomized195WW (random assignment)23 monthsNot reported78% at 5 years
Prednimustine40 monthsNot reported70% at 5 years
Interferon alpha32 monthsNot reported84% at 5 years
Ardeshna et al (2003)Randomized309cWW (random assignment)2·6 years66 months58% at 5 years; 22% at 15 years
ChlorambucilNot reported43 months57% at 5 years; 21% at 15 years
Solal-Céligny et al (2012)Prospective, observational349WW (defined as time to treatment >90 d)14 months79% at 4 years87% at 5 years
Treated low tumour burden patientsNot reported69% at 4 years88% at 5 years
Ardeshna et al (2014)Randomized463WW (random assignment)31 monthsNot reported94% at 3 years
Rituximab (induction)NRNot reported96% at 3 years
Rituximab (induction and maintenance)NRNot reported97% at 3 years
El-Galaly et al (2015)Retrospective841ActiveNot reportedNot reported78% at 5 years
WW (defined as time to treatment >90 d)33 monthsNot reported83% at 5 years (P = 0·02)
Nastoupil et al (2016)Prospective, observational1754R-CHT7 years7 years72% at 8 years
Rituximab4 years4 years67% at 8 years
WW (defined by treating physician)2·4 yearsRituximab 4 years74% at 8 years
RCVP 6·7 years
RCHOP NR

Nonetheless, some questions regarding definitions and endpoints in WW in FL remain unclear. First, given that FL is often a slow growing tumour and patients can live asymptomatically for a long time (even with active disease), progression-free survival (PFS) is an outcome of questionable value for the comparison of WW and active treatment strategies (Reagan & Friedberg, 2015). The same argument applies for time-to-treatment when comparing WW with any other treatment as it compares delaying therapy (Solal-Céligny et al, 2012) with an active treatment upfront, and patients in the WW group can receive the same treatment when they become symptomatic. Nastoupil et al (2016) measured PFS from first active lymphoma treatment, a more meaningful and informative endpoint, similarly to a previous study (Solal-Céligny et al, 2012) which determined freedom from treatment failure (in which the first treatment in the WW group was not considered an event). There were no differences between expectant management and immediate treatment in either of these outcomes. The results of these studies, together with a recent series from the Danish registry (El-Galaly et al, 2015), which also described a similar OS on comparing WW and immediate treatment, are very important and legitimize the role of WW despite the plethora of currently available treatments for FL.

A less studied but also important issue is the definition of WW. In randomized trials, patients with low tumour burden are assigned to WW or active treatment regardless of how long they ultimately remain free from treatment. Conversely, WW has also been defined in other series (observational and/or retrospective) by the length of time between diagnosis and treatment, often 90 d (Friedberg et al, 2009; Solal-Céligny et al, 2012; El-Galaly et al, 2015). Even within the framework of the NLCS, both definitions, i.e., intention of the treating physician and time from diagnosis to treatment, have been employed (Friedberg et al, 2009; Wagner-Johnston et al, 2015; Nastoupil et al, 2016). However, these two populations may differ significantly. For instance, while in randomized trials all patients have a low burden of disease; this is not the case in retrospective studies that define WW according to time from diagnosis to treatment (El-Galaly et al, 2015; Nastoupil et al, 2016). Along the same line, some randomized trials have used poor prognostic features, such as elevated serum lactate dehydrogenase (LDH) as exclusion criteria for WW (Ardeshna et al, 2014), unlike retrospective studies. Importantly, elevated serum LDH has been shown to predict a shorter time to treatment (Solal-Céligny et al, 2012; El-Galaly et al, 2015) and histological transformation, suggesting that different definitions of WW could also partly explain the disparities observed in the risk of histological transformation in WW patients versus active treatment (Montoto & Fitzgibbon, 2011).

Admittedly, the information available (such as the intent of the treating physician) in retrospective series is often limited, but we believe that it is essential that different definitions of WW be kept in mind when comparing the results obtained in different series. At the same time, the fact that, despite different populations, no study has found an increased OS with any active treatment strategy over WW seems to make WW a very valid therapeutic option for asymptomatic patients with FL in the immunochemotherapy era.

Author contributions

M.S., J.S. and J.R. drafted and revised the manuscript.

Ancillary