REGIONAL ANESTHESIA AND ACUTE PAIN
ORIGINAL ARTICLE
Electron Microscopy of Dural and Arachnoid Disruptions
After Subarachnoid Block
Miguel Angel Reina, MD, PhD,* Anna Puigdellvol-Snchez, MD, PhD,
Stephen P. Gatt, MD, FANZCA, FRCA, FCICM, FACHSM,|| Jos De Andrs, MD, PhD, FIPP, EDRA,**
Alberto Prats-Galino, MD, PhD, and Andr van Zundert, MD, PhD, FRCA, FANZCA
Background: It has been customary to attribute postdural puncture head-
ache (PDPH) incidence and severity to size and nature of the dural hole
produced during major neuraxial blockade or diagnostic dural puncture.
Needle orientation in relation to the direction of dural fibers was thought
to be of importance because of the propensity for horizontal bevel place-
ment to cause cutting rather than splitting of the dural fibers.
Methods: In vitro punctures of stringently quality-controlled human
dural sac specimens were obtained with 27-gauge (27G) Whitacre needle
(n = 33), with 29G Quincke used parallel to the spinal axis (n = 30), and
with 29G Quincke in perpendicular approach (n = 40). The samples were
studied with a scanning electron microscope, and the perimeter, appear-
ance, and area (%) of the lesion were calculated.
Results: When using small 27G to 29G needles, neither needle tip char-
acteristics nor needle orientation had a substantial bearing on the damage to
dural fibers in the dural lesion. Of ultimate importance was the characteris-
tic and size of the hole in the arachnoid. Arachnoid layer lesions produced
by different types of spinal needles were not markedly different.
Conclusions: Accepted theories of the etiology of PDPH need to be re-
vised. This article marks the first time that arachnoid layer damage has
been quantified. Dural fibers tend to have sufficient memory to close
back the hole created by a spinal needle, whereas arachnoid has diminished
capacity to do so. The pathogenesis of PDPH and its resolution algorithm
are a far more complex process that involves many more stages of devel-
opment than hitherto imagined.
(Reg Anesth Pain Med 2017;42: 0000)
I ntracranial hypotension due to cerebrospinal fluid (CSF) leak-
age with resultant traction on pain-sensitive meningeal struc-
tures is commonly proposed to explain postdural puncture
From the *Department of Clinical Medical Sciences, CEU San Pablo University
School of Medicine; and Department of Anesthesiology, Madrid-Monteprncipe
University Hospital, Madrid; Laboratory of Surgical NeuroAnatomy, Human
Anatomy and Embryology Unit, Faculty of Medicine, Universitat de Barcelona,
Barcelona; and Antn Borja Primary Care Centre, Terrassa Health Consor-
tium, Rub, Spain; ||Department of Anesthesia, Royal Hospital for Women &
Prince of Wales & Sydney Children's Hospitals, Randwick & University of
New South Wales, Kensington, Sydney, New South Wales, Australia; **De-
partment of Anesthesia Critical care and Pain Management, General University
Hospital, Valencia and Department of Surgical Specialties, School of Medicine,
Valencia University, Valencia; and Laboratory of Surgical NeuroAnatomy,
Human Anatomy and Embryology Unit, Faculty of Medicine, Universitat de
Barcelona, Barcelona, Spain; and Department of Anesthesiology & Perioper-
ative Medicine, The University of Queensland & Royal Brisbane & Women's
Hospital, Brisbane, Queensland, Australia.
Accepted for publication May 7, 2017.
Address correspondence to: Miguel Angel Reina, MD, PhD, Department of
Clinical Medical Sciences, CEU San Pablo University School of Medicine,
Valmojado, 95 1B, (Postal code 28047) Madrid, Spain
(email: miguelangel@perticone.e.telefonica.net).
The authors declare no conflict of interest.
This work was supported in part by the Investigation Funds, Ministry of Health
of Spain (project 98/0628).
Copyright 2017 by American Society of Regional Anesthesia and Pain
Medicine
ISSN: 1098-7339
DOI: 10.1097/AAP.0000000000000667
Regional Anesthesia and Pain Medicine Volume 42, Number 6, November-December 2017
Copyright 2017 American Society of Regional Anesthesia and Pain Medicine. Unauthorized reproduction of this article is prohibited.
headache (PDPH).13 There is a direct relationship between size
of the dural puncture hole and resulting CSF leak and between ex-
ternal diameter of the needle and incidence of PDPH. A larger-
size dura-arachnoid puncture produces more CSF leakage so that
the incidence of hypotension and PDPH is higher. Larger needles
produce more PDPH, and conversely, a decrease in PDPH is ap-
preciated when smaller needles are used.46 It has been accepted
that larger-size dural punctures are associated with a higher inci-
dence of PDPH.46
These clinical findings led to several hypotheses and theories
aimed at explaining mechanisms for the CSF leakage and subse-
quent headaches: (i) most theories were based on the classic de-
scription of a dura mater composed of longitudinal fibers running
parallel to each other oriented in the same axis as the spinal cord.
Many believed that the dura mater healing process was exclusively
responsible for stopping the leakage of CSF after dural puncture79;
(ii) for decades, it was recommended that, during lumbar punc-
tures with Quincke-type needles, the bevel be oriented parallel
to the axis of the spine and the dural fibers. These recommenda-
tions assumed that dural lesions produced by the tip of the needles
would be smaller9 because parallel orientation of the needle tip
bevel would allow separation of dural fibers, whereas introducing
the bevel perpendicular to the axis of the spinal cord and dural fi-
bers would cause larger puncture lesions by sectioning the dural
fibers911; (iii) it was in the 1990s that smaller spinal needles be-
came the standard and pencil-point needle-tip designs were fa-
vored. Studies demonstrated that the incidence of PDPH was
reduced with pencil pointtype needles in comparison with Quincke-
type needles. The hypothesis of choice was that these needle tips
allow separation rather than cutting of dural fibers and therefore
produced smaller dural lesions.1214 These assumptions, however,
are neither substantiated nor supported by morphological evidence.
Over the years, a cyclical argument developed in which sup-
port for the popular mechanical hypotheses for dural lesion pro-
duction combined with a low incidence of PDPH with smaller
needles fueled entrenchment of anesthetists' views with unfounded
assumptions being made about histological changes at the site of
dural lesions.8 Consequently, few contributions to the exact mor-
phology of dural lesions produced by spinal needles were made.
In the recent past, our group performed several experiments
and have been accumulating evidence on human meninges (spinal
dura mater and arachnoid layer)1521 aimed at observing and
documenting dura-arachnoid lesions produced by 25-gauge (25G)
Quincke and Whitacre needles and 22G Quincke needles.22,23
In this study, our aims were (a) to study the morphology and
size of dural sac lesions with both 27G Whitacre and 29G
Quincke spinal needles, (b) to examine the microanatomy of the
human dural sac, (c) including the arachnoid layer, and (d) to view
appearance of lesions in order to confirm whether they complied
with our impressions of lesions produced by larger-diameter
needles22,23 and (e) whether the diameter of lesions correlated
with the needles used. Finally, we studied whether ( f ) the lesion
diameter produced with 29G spinal needles was altered when
1
Reina et al Regional Anesthesia and Pain Medicine Volume 42, Number 6, November-December 2017
needle bevel was oriented vertically as opposed to the tip being
oriented horizontally in relation to the spinal axis.
METHODS
After ethics committee approval (CEIC code 13.11.543-GHM),
images were analyzed following in vitro punctures of optimally proc-
essed dural sac specimens using single-use, new 27G Whitacre
(n = 33), 29G Quincke used parallel to the spinal axis (n = 30),
and 29G Quincke needles in perpendicular approach (n = 40).
Images obtained were grouped into 6 subgroups: group 1: le-
sions produced by 27G Whitacre needle studied on the dural surface;
group 2: lesions produced by 27G Whitacre needle observed on
arachnoid surface; group 3: 29G Quincke with bevel alignment par-
allel to axis lesions observed on dural surface; group 4: as in group
3 but observed on arachnoid surface; group 5: 29G Quincke with
bevel alignment perpendicular to axis lesions observed on dural
surface; group 6: as in group 5 observed on arachnoid surface.
In Vitro Methodology
Spinal cord and dural sac were dissected and removed from
the eighth thoracic (T8) vertebra to the fifth lumbar (L5) vertebra
from 8 fresh cadavers, 23 to 60 years of age. Nerve roots crossing
the dural sac were removed to leave the dural sac isolated from the
dura and the arachnoid mater.
To retain the anatomical spinal cylindrical structure and to re-
move irregularities within the membranes and retain physiological
radial forces, the dural sac was mounted onto 3 aluminum rods
covered with latex. The dura mater was kept in a stable physiolog-
ical environment by continuously spraying a saline solution onto
the sacs to prevent histological deterioration of the samples.
All punctures were done at 90 degrees to the surface of the
dura. Needles (BD, Madrid, Spain) were never reused. Punctures
were performed and examined at random. All in vitro punctures
were performed by the same researcher (M.A.R.), with fixation
and processing of samples and scanning electron microscope
(E/M) image capture being entrusted to 2 blinded technicians.
Only after the measurements were completed were results
assigned into each group.
In vitro punctures imitated the identical conditions in clinical
practice with regard to speed of advancement and depth of needle
puncture. In all samples, the dural sac was completely traversed by
the needle and extended a few millimeters beyond the needle tip.
Approximately 30 seconds after puncture, the needle was care-
fully removed, taking care not to cause additional damage to the
tissue sample.
Approximately 20 minutes after the punctures, all samples
were processed and later studied under a scanning E/M. Some proc-
essed samples were studied on the outer epidural surface and others
on the inner intrathecal surface with a JEOL JSM 6400 Scanning
Electron Microscope (JEOL Corporation Ltd, Tokyo, Japan).
Adobe Photoshop Image J version 1.46a (http://rsb.info.nih.
gov/ij/) software was used to homogenize magnifications among
images and to allow measurement of perimeter and area of dural
and arachnoid lesions. The characteristics of the edges of the le-
sions on the arachnoid layer and the external side of the dural lam-
ina were observed, and the perimeter and area percentages of the
lesion characteristic when compared with the perimeter and the
area of the needle used were calculated.
The 27G Whitacre needle was chosen because it remains one
of the best and most versatile spinal needles commonly used in
neuraxial regional anesthesia and the 29G Quincke needle because
it is the finest and thinnest spinal needle made and used in
clinical practice.
2 2017 American Society of Regional Anesthesia and Pain Medicine
Copyright 2017 American Society of Regional Anesthesia and Pain Medicine. Unauthorized reproduction of this article is prohibited.
Statistical Analysis
Shapiro-Wilk test was used to test the normality in small sam-
ples before using the t test. When normality was not confirmed,
Mann-Whitney U test was used for nonparametric comparisons.
Values are expressed as mean of dural and arachnoid areas and perim-
eters and percentages regarding equivalent characteristics of the cor-
responding needles. P < 0.05 was considered statistically significant.
RESULTS
From a total of 120 punctures performed, 49 dural and 54 arach-
noid lesions were analyzed: group 1: 19 images, group 2: 14 images,
group 3: 16 images, group 4: 14 images, group 5: 14 images, and
group 6: 26 images. The rest of the samples (n = 17) were discarded
because of artifacts during laboratory processing or difficulty localizing
peripherally located lesions under the E/M or because holes created on
a specimen were too close to a margin causing uneven shrinkage.
It was only possible to study 1 surface of specimens (external
or internal) from each puncture because the probe used in the mi-
croscope had 1 free surface of tissue that could be studied with the
other surface lying in contact with an adhesive.
The results of the area and the perimeter of dural and arach-
noid lesions and needles used are shown in Table 1. Dural and
arachnoid lesions produced by 27G Whitacre and 29G Quincke
needles are shown in Figures 1 to 4. Dura-arachnoid complex
puncture produces, initially, a tent-like indentation followed by
piercing of the dura and arachnoid mater. Insertion of a single-
use, beveled spinal Quincke needle produces less tent-like effect
(tenting) than a noncutting pencil-point Whitacre needle. The
tip of pencil-point needles causes more distension of the dural
sac before the tissue fabric is broken. We observed that the super-
ficially located external dural laminae remained partially open,
whereas deeper dural laminae were most often fully closed
(Figs. 1 and 3). On the other hand, we observed, in many samples,
that the internal arachnoid layer was partially open, whereas others
were better approximated so as to appear almost closed (Figs. 2
and 4). In samples with the arachnoid lesion partially open, we
would often observe, deeper down through the arachnoid layer's
hole, the dural laminae partially or fully closed. The lesions were
always made up of a dural structural lesion underpinning the
arachnoid lesion. In this context, closed applies to lesion edge
approximation rather than physical sealing of the hole flaps.
Observation of the lesions themselves using the scanning
E/M produces 2 very distinctly different pictures:
(1) Quincke: The bevels on Quincke-type spinal needles pos-
sess 2 distinct components: (a) a distal beveled tip that pro-
duces a small cutting lesion of the dura-arachnoid complex
and (b) sharp-edged shoulders. As the tip advances, the
sharp edges cut and distend the dura, favoring clean pro-
gression through the complex. Quincke spinal needles pro-
duce a precise, controlled cut, limiting damage of dura
mater and arachnoid layer. The tip of the needle folds
cut tissue inside the dura so that dural and arachnoid lesions
produced by Quincke needles show a crescent shape re-
sembling the letters U or V, not dissimilar to the lid of
an opened can. The edges are clean cut, and tissue loss
is minimized. Only a thin orifice remains open in lesions
caused by beveled cutting spinal needle tips.
(2) Whitacre: Pencil-point tips produce greater tearing of tissue
margins and folding at the edges of the lesion. The edges of
dural and arachnoid lesions caused by pencil-point needles
are jagged and irregular. The tissue margins are fragmented
because the process of needle advancement is one of blunt
tearing and avulsion.
Regional Anesthesia and Pain Medicine Volume 42, Number 6, November-December 2017 Arachnoid and Dura Mater Lesions

TABLE 1. Area (in mm2), Perimeter (in mm), and Percentage Respect Same Data of Needle Used on Dural and Arachnoid Lesions (in %)

29G Quincke Needle: 29G Quincke Needle:

Type 27G Whitacre Needle Parallel Bevel Perpendicular Bevel
Area, mm2 0.1 0.08 0.08
Spinal Needle Diameter/perimeter, mm 0.40/1.26 0.33/1.04 0.33/1.04
Group 1 Group 3 Group 5
Dura mater n 19 16 14
Area SD, mm2 0.018 0.009 0.015 0.006 0.015 0.007
Percentage lesion area/needle area, 15.2 18.3 19
Perimeter SD, mm 0.848 0.195 0.771 0.107 0.777 0.163
% Lesion perimeter/needle perimeter 67.3 74.1 74.7
Group 2 Group 4 Group 6
Arachnoid mater n 14 14 26
Area SD, mm2 0.028 0.015 0.016 0.014 0.014 0.011
% Lesion area/needle area 23 19.9 17.6
Perimeter SD, mm 0.914 0.195 0.769 0.189 0.738 0.212
% Lesion perimeter/needle perimeter 72.5 74.0 71
Data are presented as mean SD or number (%).

There were significant differences between the dural and
arachnoid lesion area in 27G Whitacre (P = 0.029) but not in pe-
rimeter length (P = 0.345).
There were no significant differences between the parallel
and perpendicular bevel approach using the same 29G Quincke
needle type on dural surface lesions (group 3 vs group 5: dural area,
P = 0.880; dural perimeter, P = 0.910) or on arachnoid surface le-
sions (group 4 vs group 6: arachnoid area, P = 0.940; arachnoid pe-
rimeter, P = 0.640). There were no significant differences between
the dural and arachnoid lesions in 29G Quincke needles using ei-
ther parallel orientation (area, P = 0.142; perimeter, P = 0.377) or
perpendicular alignment (area, P = 0.361; perimeter, P = 0.305).
There were significant differences between the size (area) of
lesions produced by 27G Whitacre needle and overall 29G Quincke
needle lesions on the arachnoid surface (area, P = 0.001; perimeter,
P = 0.011) but not on dural surface (area, P = 0.181; perimeter,
P = 0.234). Such difference in arachnoid layer area between 27G
Whitacre needles and 29G Quincke needles was confirmed in both
parallel alignments for area (P = 0.008) and perimeter (P = 0.024)
and in perpendicular alignments in both area (P = 0.001) and perim-
eter (0.039). Dural lesions showed no differences between 27G
Whitacre and 29G Quincke needles with either parallel alignment
(area, P = 0.182; perimeter, P = 0.271) or with perpendicular orien-
tation (area, P = 0.377; perimeter, P = 0.280).
DISCUSSION
This study demonstrates that (1) accepted theories of the eti-
ology of PDPH need to be revised; (2) with 27G or smaller-size
needles, the lesions produced by different needles are small
there is no difference between needle orientation in the nature of
the lesions produced; (3) it is the arachnoid damage that contrib-
utes primarily to PDPH characteristics; (4) dural fibers tend to
have sufficient memory to close back the hole created by a spinal
needle, whereas arachnoid has diminished capacity to do so;
(5) spurious theories regarding dural fiber orientation are
incorrectthe distribution of fibers is not unidirectional but
randomly laid out in concentric, independent layers, and theories
based on fiber splitting rather than cutting are a little more than
myth; (6) pencil-point needle injury wounds are avulsion
lesions not splitting lesions; (7) once needles of less than 27G
are selected by clinicians, other needle characteristics (eg, shaft ri-
gidity, which determines the extent of deviation from the direc-
tional axis of entry; internal area of the hollow needle, which
determines speed of CSF flashback) become more important pri-
orities when compared with needle tip style, needle shoulder
bevels, or tip orientation relative to spinal axis; (8) according to
our studies, in vitro, the lesion produced by the 29G Quincke nee-
dle is smaller than that produced by the 27G Whitacre needle. Ad-
ditional factors applicable to clinical practice may be evaluated in
future studies; (9) even though PDPH is not recognized as a sig-
nificant complication of spinal sac puncture with fine needles,
the pathogenesis of PDPH and its resolution algorithm are a far
more complex process that involves many more stages of devel-
opment than hitherto imagined; (10) the arachnoid barrier and
dura superstructure frame functions are emphasized, and (11)
an improved, uniform, standardized, and universal electron mi-
croscopy methodology for the management and measurement of
dura-arachnoid complex lesions has been used.
The dural hole obtained after a lumbar puncture consists of
2 components: (a) a dural lesion obtained by piercing several lam-
inae of the dura mater and (b) an arachnoid layer puncture.
As a Quincke needle tip is advanced, the jagged fragments at
the edge of the lesions tend to fold inward, after all dural and
arachnoid layers are severed, and the edges of the lesion are
displaced away from the needle tip. As soon as the spinal needle
is withdrawn, the edges of the lesion tend to retract because of
the viscoelastic properties of the affected dura mater. This may fa-
vor the return of the ragged edges to their original position, has-
tening the closure of the dural sac lesion.
Pencil-point spinal needles of Whitacre style have noncutting
tips and tend to displace the dural sac before piercing the dural
laminae during a lumbar puncture. Lack of sharp beveled tips
produces a larger tenting indentation before the dura-
arachnoid penetration.
Closure of the lesion produced by either needle is a complex
phenomenon with many mechanisms at play. One could hypothe-
size that the more pronounced avulsion of fibers produced by
pencil-point needles, when compared with the cleaner cut made
by Quincke-type needles, could increase the local inflammatory

2017 American Society of Regional Anesthesia and Pain Medicine 3

Copyright 2017 American Society of Regional Anesthesia and Pain Medicine. Unauthorized reproduction of this article is prohibited.
Reina et al Regional Anesthesia and Pain Medicine Volume 42, Number 6, November-December 2017

FIGURE 1. A, Dural sac lesion produced by a 27G Whitacre needle seen from the epidural surface of the dura mater, that is, the most external
surface of human dural sac showing the residual aperture of the lesion following needle withdrawal. The deeper dural laminae are partially
closed, that is, in closer anatomical approximation. Scanning electron microscopy, original magnification 200; index ruler bar, 100 m. B,
Colors enable identification of respective structures to demonstrate how partially closed outer, more superficial dural laminae overly
almost closed deeper dural laminae. Superimposed over the E/M image is the outer circumferential perimeter of the respective needle used.
The gray tissue represents the enclosing, fully recovered outermost dural lamina after the needle has been withdrawn.

reaction and associated edema in vivo, resulting in faster closure of
the traumatic lesion.22 This hypothesis has never been confirmed.
It should be remembered that our study results reflect a snap-
shot of closure of the dural and arachnoid holes caused by the 2
different types of fine needles under very specific conditions and
may therefore not reflect the entire process of damage and repair.
For example, (1) examination of the closure process was arrested
within approximately the first 20 minutes. Once the tissue samples
were fixed, viscoelastic properties no longer apply so that the
size and shape of the lesions could no longer change spontane-
ously; (2) once the samples were extracted, the effects of pressure
from CSF pressing outward across the dural sac lesion and CSF
flow past the edges of the lesion are removed. This could have
an effect on the final rearrangement of the lesion edges around
the zone of closure so that the in vitro conditions may not mimic
entirely what actually happens in real time.
If these limitations are recognized and accepted and keeping in
mind that observations are restricted to the structure of early lesions in
the dura-arachnoid during lumbar spinal sac puncture with small-
gauge needles, it is interesting to note that lesions observed are not
circular and suffer a rearrangement process with approximation of
edges. Approximation of edges of the puncture hole may also occur
in vivo because of the dural viscoelastic properties. Indeed, each nee-
dle type and size exert a different stretching force or tenting effect.

4 2017 American Society of Regional Anesthesia and Pain Medicine

Copyright 2017 American Society of Regional Anesthesia and Pain Medicine. Unauthorized reproduction of this article is prohibited.
Regional Anesthesia and Pain Medicine Volume 42, Number 6, November-December 2017 Arachnoid and Dura Mater Lesions

FIGURE 2. A, Similar arachnoid and dural lesion produced by a 27G Whitacre needle to that in Figure 1 viewed from the intrathecal, internal
surface of the human dural sac showing both the arachnoid layer and the most internal dural laminae of dura mater partially closed through
the orifice of the arachnoid lesion. Scanning electron microscopy, original magnification 200; bar = 100 m. B, Colors enable identification of
the (partially apposed) inner dural laminae, that is, those closest to the arachnoid layer, and the (almost closed) deeper dural laminae. The
circumference of the lesion depends on the external diameter of the needle used. The gray tissue at the periphery of the image represents the
stretched innermost arachnoid lamina, which has recoiled back into position after the needle has been withdrawn. Of note in this image is
that the length or extent of the lesion exceeds the hypothetical diameter of the superimposed needle's outer perimeter. This is explained by the
larger circular hole boundary or edge having collapsed into a more linear array after recoil and retraction.

Electron microscopy examination of the epidural or outer
surface of the dural sac shows extent of closure of dural laminae
within the dura mater, whereas observation of the intrathecal or
inner surface of the dural sac permits study of the nature and
early resealing of the hole in the arachnoid layer (Figs. 14).
Furthermore, E/M examination through the inner arachnoid layer
allows observation of partial or even complete closure of internal
dural laminae (Figs. 2 and 4). A different degree of closure of
each dural sheet lesion is possible because the dura mater is made
up of multiple concentric dural laminae.20 Each lamina acts as an
independent unit so that the mode and extent of closure, response
to trauma, mechanisms for healing, and damage sustained at
puncture are both similar and different for each structural and
functional unit within the dura mater. These observations cast a
novel light onto understanding how rearrangement of the edges
of dural lesion affects the lesion's area parameter rather than
simple measurement of the corresponding perimeters, that is,
that both size of hole and perimeter characteristics, especially on
the arachnoid inner surface, are important, we surmise, in
establishing rate of dural CSF leak and lesion closure time. In
contrast to the dura mater, the arachnoid layer is formed of
arachnoid cells firmly attached by specialized membrane junctions

2017 American Society of Regional Anesthesia and Pain Medicine 5

Copyright 2017 American Society of Regional Anesthesia and Pain Medicine. Unauthorized reproduction of this article is prohibited.
Reina et al Regional Anesthesia and Pain Medicine Volume 42, Number 6, November-December 2017

FIGURE 3. A, Representative crescentic U shape of one of the dural lesions at 200 magnification scanning E/M magnification produced by
a 29G Quincke needle on the external, epidural surface of human dural sac showing the characteristics of the aperture of the lesion on the
most external dural laminae of the dura mater. Deeper into the lesion, dural laminae can be seen partially closed (index bar = 100 m). B, Color
labeling allows identification of outer (open or partially closed) superficial dural laminae and deeper (almost closed) dural laminae. The
circumference of the lesion is proportional to the external diameter of the needle used. Open = ability to see straight across the lesion into the
intrathecal space; closed = inability to see through all layers of the lesion; partially or almost closed = degree of approximation of lesion
circumferential structures.

(tight junctions and desmosomes) forming the major meningeal perimeter characteristic of the arachnoid layer lesion that is of
barrier of the dural sac. Images from our previous researches20,21 even greater importance.
show such differences (Fig. 5). The arachnoid layer has no The reduced hole area and length of perimeter in the dural
directionality of its components,19,21 so that it is now the laminae in 29G Quincke compared with 27G Whitacre needles in

6 2017 American Society of Regional Anesthesia and Pain Medicine

Copyright 2017 American Society of Regional Anesthesia and Pain Medicine. Unauthorized reproduction of this article is prohibited.
Regional Anesthesia and Pain Medicine Volume 42, Number 6, November-December 2017 Arachnoid and Dura Mater Lesions

FIGURE 4. Arachnoid surface view of the human dural sac lesion produced by a 29G Quincke needle. A, Internal, intrathecal surface showing
the open arachnoid lesion (produced on the arachnoid layer) with the broken collagen fibers of the edge virtually in apposition. Scanning
E/M, original magnification 200; bar = 100 m. B, A with coloration permitting identification of the arachnoid layer, the (partially closed)
inner dural laminae immediately underlying the arachnoid layer, and some (almost closed) deeper dural laminae. Gray tissue in this image
represents the arachnoid surface in recoil position.

percentages are almost proportional to the diameter of needle used
(Table 1). With small needles, the size of the lesion depends more on
the diameter of the needle than on the design of the tip of the needle.
In clinical practice, Whitacre needles enjoy a marginally re-
duced incidence of PDPH, but they cannot be considered to be
nontraumatic needles; the edges of the lesions produced are irreg-
ular when compared with the smooth margins of lesions caused by
the sharp, beveled cutting Quincke needles.
A reduced perimeter length and area (size) of the lesion are
produced by viscoelasticity, whereas perimeter length is affected
mainly by tissue retraction; as it partly recovers its original position,
the area is influenced by additional factors such as CSF pressure.
Importantly, as shown in Table 1, it was observed that the size of
dural lesion was reduced to 20% of the original hole size; that is,
that dura mater recovered 80% of its distended, puncture position.
Interestingly, all the dural lesion images suggest that fiber
damage, rather than fiber separation, occurs with spinal needles.
This questions the commonly held intuitive belief that dural fiber
separation, rather than fiber damage, occurs with pencil-point and
bullet-head puncture needles. Because we demonstrate an irregu-
lar border of the lesion perimeter instead of a linear split, as would
be expected if fiber separation were to occur, we conclude that at-
tributing diminished PDPH to Whitacre pencil-point needles is
probably only 1 aspect, probably rather minor, of the reasons for
the desirability of using these needles preferentially. Dural fibers
have neither a longitudinal distribution nor a parallel orientation;
the fibers are randomly distributed.20,22 Favoring pencil-point spi-
nal needles on the basis that they are less traumatic is not sup-
ported by our findings.
Likewise, the concept of fiber separation utilized until now
to recommend aligning the bevel of Quincke-type needles in the
axial plane is also not supported by this study.22,23
The 29G and 27G needles studied have external diameters
ranging from 330 to 400 m. These dimensions are orders of

2017 American Society of Regional Anesthesia and Pain Medicine 7

Copyright 2017 American Society of Regional Anesthesia and Pain Medicine. Unauthorized reproduction of this article is prohibited.
Reina et al Regional Anesthesia and Pain Medicine Volume 42, Number 6, November-December 2017
FIGURE 5. Partial-thickness cross section of human arachnoid layer
with its tight junctions (A) and a separate partial-thickness view of
human dura mater (B). The arrow shows the direction of diffusion
between epidural space and arachnoid space in both directions.
The location of the blue color between arachnoid cells in A
represents the barrier to transcellular passage across the
arachnoid layer. The blue color among the collagen and elastic fibers
in B depicts the relatively easier passage across the dura mater.
Transmission electron microscopy, original magnification 30,000
(A), 20,000 (B). A and B were reproduced with permission from
M.A.R. Arach indicates arachnoid cell; Coll, collagen fiber; Elast,
elastic fibers.
magnitude larger than collagen (0.1-m diameter) or elastin
(2-m diameter) fiber ultrastructure within the dura mater.17,20
Such diameter difference, together with the shape of the observed
lesions, suggests that fiber avulsion, rather than fiber separation,
occurs. No needle, regardless of design type, can be considered
nontraumatic or atraumatic. from 27G to 29G, and area is a squared function of radius by the
When it comes to arachnoid layer ultrastructure, alignment of
bevel of Quincke-type needles is even more irrelevant because
arachnoid cells have neither longitudinal distribution nor parallel
orientation and are much smaller than the elastin fibers.19,21 This
study confirms that there were no significant differences in the
size (area and perimeter) of lesions analyzed when the lumbar
puncture was performed with the bevel of the 29G Quincke-type
needle oriented parallel or perpendicular to the spinal axis.
Although counterintuitive, the finding that the dura-arachnoid
lesions cannot be related to any fiber splitting mechanism applies,
to varying degrees, to other needle sizes. Figure 6 compares the
lesions produced by 27G Whitacre and 29G Quincke needles with
those produced by 22G and 25G Quincke and 25G Whitacre
8 2017 American Society of Regional Anesthesia and Pain Medicine
Copyright 2017 American Society of Regional Anesthesia and Pain Medicine. Unauthorized reproduction of this article is prohibited.
needles, captured at the same magnification, from our previously
published articles.22,23
Although superficially, there may seem to be a discrepancy
between this study's E/M microanatomical findings and in vivo
clinical experience; deeper examination of the clinical studies
shows wide variability of results. To cite 2 simple examples where
clinical practice does not mimic laboratory conditions, (a) each of
our E/M images is for a single clean puncture, whereas in clinical
practice, it is sometimes necessary to perform several puncture at-
tempts before CSF is obtained. This watering-can multiple-hole
array may exhibit totally different characteristics to a single
hole; (b) our brand new, single-use needles were used once and
discarded; in clinical practice, if the first attempt fails, the needle
is technically blunt from deformation of the tip especially if os-
seous structures, for example, spinous process or vertebral lamina,
are encountered. This effect is magnified in situations (eg, coun-
tries with limited resources) where spinal needles are resterilized
and reused. Tip deformation is worst with Quincke-type needles
and is virtually minimal with Whitacre pencil-point needles.24
Some insertion techniques could reduce vertebra impingement
and consequent needle deformation.25
In our studies, we tested only single punctures using high-
quality, new, pristine propriety needles of original design. Blunt,
deformed, poor-quality, multiple-use, or poorly designed needles
in multipuncture situations were not looked atm but we surmise
that in such studies wounds will be larger, more jagged, and less
likely to close and will only magnify our findings and render
our conclusions even more relevant.
On the other hand, our in vitro studies were performed on
specimens of isolated dural sacs that were not subjected to tensile
forces that occur during changes of posture and position.
The etiology and pathogenesis of PDPH, especially, although
less commonly encountered, with fine-gauge spinal needles, are
not fully understood. It cannot be solely explained by the type
and size of needle used. Many other factors contribute, so that
the process initiated by the dural and arachnoid puncture leads
to a complex interaction of factors and healing processes. The
dural and arachnoid layers have entirely different structure and
function and heal differently.26,27
This study analyzes lumbar punctures in ideal conditions
with direct visualization of location of puncture using undamaged,
undeformed needles in a single piercing at the first attempt. Nev-
ertheless, these data contribute to new ideas and hypotheses, some
of which do not support the currently espoused putative mecha-
nisms in current clinical studies.
The Whitacre 27G suffers from other design deficiencies, as
does the Quincke 29G. These extremely thin needles with low in-
ternal cross-sectional area are more likely to be deflected laterally
away from the midline and have poor flashback time (delayed
exit of CSF into the hub even when the distal needle aperture is in-
trathecal). Because internal radius drops dramatically as one goes
equation A = r2, in practice most practitioners will opt for a 27G
as opposed to a 29G needle. This study confirms that lesions pro-
duced by 27G Whitacre needles under laboratory conditions are
greater than lesions produced by 29G needles, but theoretical ad-
vantages (less discomfort, lower PDPH incidence) are outweighed
by other disadvantages (deflection, slow flashback, multiple at-
tempts, failed puncture); that is, the risk-benefit ratio of 29G
needles becomes vanishingly low.
A very relevant aspect of these studies is that the lesion pro-
duced within the arachnoid layer is at least as important as the dura
mater hole on subsequent genesis of CSF loss because the dura,
unlike the arachnoid, lacks cellular tight junctions. Closure of
the dura mater whose fibers lack spatial orientation would not,
Regional Anesthesia and Pain Medicine Volume 42, Number 6, November-December 2017 Arachnoid and Dura Mater Lesions

FIGURE 6. Comparative images of lesions produced by needles of different sizes all viewed from the arachnoid surface of human dural sac.
Images of lesions produced by 25G Whitacre, 25G Quincke, and 22G Quincke needles obtained in previous studies of our team are
included, with permission,22,24 for purposes of comparison of size (area), appearance, and periphery (circumference) characteristics. All
images: scanning electron microscopy at magnification 200.

alone, prevent further CSF leakage until the subjacent arachnoid
layer had been completely restored. This phenomenon does not di-
minish the importance of the dural layer in the healing process be-
cause dura provides the lattice superstructure along which the
arachnoid can seal the gap.
There is no benefit in adopting a specific needle-tip type or
alignment of sharp beveled tips given the rectangular shape of
arachnoid cells and the absence of a specific spatial orientation.
In summary, lesions produced by fine spinal needles are
small. While the dura mater's viscoelastic property has a much
better initial capacity to reduce hole size and CSF leak than the
arachnoid layer, it is synergistic arachnoid healing that completely
stops CSF leakage. This study underpins the importance of further
studying the arachnoid layer in the etiology of PDPH. While this
study does not give a definitive answer as to the origin and pro-
gression of PDPH, the results contribute to a better understanding
of its etiology and dismiss some myths relating to possible con-
tainment and harm-reduction strategies.
ACKNOWLEDGMENT
The authors thank Alfredo Fernndez Larios, Alfonso Rodrguez
Muoz, and Ana Vicente Montaa, from the National Center for
Electron Microscopy ICTS, Madrid, for their technical assistance
in the use of scanning E/M.
REFERENCES
1. Benito-Len J, Reina MA, Alvarez-Linera J. Intracranial hypotension
sndrome. Neurologia. 2001;16:418426.
2. Longo S. Postdural puncture: implications and complications. Curr Opin
Anaesthesiol. 1999;12:271275.

2017 American Society of Regional Anesthesia and Pain Medicine 9

Copyright 2017 American Society of Regional Anesthesia and Pain Medicine. Unauthorized reproduction of this article is prohibited.
Reina et al Regional Anesthesia and Pain Medicine Volume 42, Number 6, November-December 2017
3. Van Zundert A, Goerig M. August Bier 18611949. A tribute to a
great surgeon who contributed much to the development of modern
anesthesia on the 50th anniversary of his death. Reg Anesth Pain Med.
2000;25:2633.
4. Candido KD, Stevens RA. Post-dural puncture headache: pathophysiology,
prevention and treatment. Best Pract Res Clin Anaestesiol. 2003;17:
451469.
5. Davignon KR, Dennehy KC. Update on postdural puncture headache.
Int Anesthesiol Clin. 2002;40:89102.
6. Turnbull DK, Shepherd DB. Post-dural puncture headache: pathogenesis,
prevention and treatment. Br J Anaesth. 2003;91:718729.
7. Franksson C, Gordh T. Headache after spinal anesthesia and a technique for
lessening its frequency. Acta Chir Scand. 1946;94:443454.
8. Killian H. History of local anesthesia. In: Killian H, ed. Local Anesthesia.
Barcelona, Spain: Salvat Editores; 1979:312.
9. Mihic DN. Postspinal headache and relationship of needle bevel to
longitudinal dural fibers. Reg Anesth. 1985;10:7681.
10. Flaatten H, Thorsen T, Askeland B, et al. Puncture technique and
postural postdural puncture headache. A randomised, double-blind study
comparing transverse and parallel puncture. Acta Anaesthesiol Scand.
1998;42:12091214.
11. Greene HM. Lumbar puncture and prevention of postpuncture headache.
JAMA. 1926;86:391392.
12. Halpern S, Preston R. Postdural puncture headache and spinal needle
design. Metaanalyses. Anesthesiology. 1994;81:13761383.
13. Hart JR, Whitacre RJ. Pencil-point needle in prevention of postspinal
headache. J Am Med Assoc. 1951;147:657658.
14. Sprotte G, Schedel R, Pajunk H, Pajunk H. An atraumatic universal
needle for single-shot regional anesthesia: clinical results and a 6 year trial
in over 30,000 regional anesthesias. Reg Anaesth. 1987;10:104108.
15. Dittmann M, Reina MA, Lpez Garca A. New results in the visualization
of the spinal dura mater with scanning electron microscopy. Anaesthetist.
1998;47:409413.
16. Reina MA, Lpez-Garca A, Dittmann M, de Andrs JA. Structural
analysis of the thickness of human dura mater with scanning electron
microscopy. Rev Esp Anestesiol Reanim. 1996;43:135137.
10
Copyright 2017 American Society of Regional Anesthesia and Pain Medicine. Unauthorized reproduction of this article is prohibited.
17. Reina MA, Lpez-Garca A, Dittmann M, de Andrs JA. Analysis of the
external and internal surface of human dura mater with scanning electron
microscopy. Rev Esp Anestesiol Reanim. 1996;43:130134.
18. Reina MA, Dittmann M, Lpez Garcia A, van Zundert A. New
perspectives in the microscopic structure of human dura mater in the
dorsolumbar region. Reg Anesth. 1997;22:161166.
19. Reina MA, Prats-Galino A, Sola RG, Puigdellvol-Snchez A, Arriazu
Navarro R, de Andrs JA. Structure of the arachnoid layer of the human
spinal meninges: a barrier that regulates dural sac permeability. Rev Esp
Anestesiol Reanim. 2010;57:486492.
20. Reina MA, Lpez A, Dittmann M, de Andrs JA. Ultrastructure
of spinal dura mater. In: Reina MA, ed. Atlas of Functional Anatomy of
Regional Anesthesia and Pain Medicine. New York: Springer; 2015:
411434.
21. Reina MA, Pulido P, Garca de Sola R. Ultrastructure of spinal
arachnoid layer. In: Reina MA, ed. Atlas of Functional Anatomy of
Regional Anesthesia and Pain Medicine. New York: Springer; 2015:
435454.
22. Reina MA, de Len Casasola OA, Lpez A, de Andrs JA, Martn S, Mora
M. An in vitro study of dural lesions produced by 25-gauge Quincke and
Whitacre needles evaluated by scanning electron microscopy. Reg Anesth
Pain Med. 2000;25:393402.
23. Reina MA, Lpez A, Badorrey V, de Andrs JA, Martn S. Dura-arachnoid
lesions produced by 22 gauge Quincke spinal needles during a lumbar
puncture. J Neurol Neurosurg Psychiatry. 2004;75:893897.
24. Reina MA, de Andrs JA, Lpez A. Subarachnoid and epidural anesthesia.
In: Raj P, ed. Textbook of Regional Anesthesia. New York: Churchill
Livingstone; 2002:307324.
25. Lpez A, Reina MA. Needles in regional anesthesia. In: Reina MA, ed.
Atlas of Functional Anatomy of Regional Anesthesia and Pain Medicine.
New York: Springer; 2015:829852.
26. Puigdellvol-Snchez A, Reina MA, Sala-Blanch X, Poms-Tall J,
Prats-Galino A. Pythagoras and Cosines: the skin-dural sac distance and
optimal angles in paramedian spinal anesthesia. Clin Anat. 2016;29:
10461052.
27. Van Zundert AA, Reina MA, Lee RA. Prevention of post-dural puncture
headache (PDPH) in parturients. Contributions from experimental research.
Acta Anaesthesiol Scand. 2013;57:947949.
2017 American Society of Regional Anesthesia and Pain Medicine